CN114580732A - Urban logistics vehicle path optimization method considering dynamic efficiency of driver - Google Patents

Abstract

The invention discloses an urban logistics vehicle path optimization method considering the driver's dynamic efficiency, including determining the relationship between driver fatigue and performance: introducing a fatigue curve model, analyzing the relationship between driver fatigue and performance; establishing a vehicle path optimization model : Determine the problem objectives and constraints, determine the symbolic representation of parameters and variables, and establish a mathematical model; solve the above-mentioned models: solution coding and evaluation, initialization stage design, hiring bee stage design, follower bee stage design, scout bee stage design. The present invention can not only make the logistics enterprise plan the distribution path of the driver more reasonably under the condition of limited resources, help the enterprise to improve the distribution efficiency and save the distribution cost, but also make the path planning process take into account the interests of the driver and alleviate the driver's burden. Fatigue accumulation improves the driver's work experience. These two aspects are of great significance for improving the competitiveness of enterprises.

Description

An urban logistics vehicle routing optimization method considering driver's dynamic efficiency

technical field

The invention belongs to the technical field of logistics optimization, and relates to an urban logistics vehicle path optimization method considering the dynamic efficiency of drivers.

Background technique

In recent years, driven by the continuous growth of e-commerce, urban logistics has become an indispensable part of national life. In the urban logistics activities, the driver is the link between the enterprise and the customer, and its distribution efficiency directly affects the enterprise's distribution cost, distribution efficiency and customer satisfaction level. In order to improve the delivery efficiency of the driver, the logistics dispatcher often formulates the optimal delivery route plan before the driver leaves, but the actual delivery situation is often inconsistent with it. As the delivery process may be affected by external factors such as weather and traffic, as well as inherent characteristics such as driver fatigue, the actual driving path and time spent by drivers to complete the delivery task are often different from the plans expected by the dispatcher when formulating the plan. big difference.

Up to now, researchers have paid more attention to the influence of external factors on the optimal delivery route, ignoring the influence of the intrinsic characteristics of drivers, especially the important factor of driver fatigue. In today's high-intensity working environment, the fatigue generated during the delivery process will inevitably affect the driver's performance, resulting in a reduction in the driver's delivery efficiency. Therefore, how to effectively optimize the distribution plan and provide better distribution services for customers under the premise of considering the dynamic performance of drivers caused by fatigue has become an urgent problem to be solved.

The distribution optimization problems of express delivery, takeaway, and fresh food in urban logistics can be collectively referred to as vehicle routing problems. The vehicle routing problem is a typical NP-hard problem, and the current research on solving methods mainly focuses on the meta-heuristic method. Among them, the artificial bee colony algorithm, as a classic meta-heuristic algorithm, was first formally proposed by Karaboga in 2005. The algorithm simulates the swarm intelligent behavior of bees collecting honey in nature. A good source of honey. In practical optimization problems, the process of bees searching for the optimal nectar source is the process of finding the optimal solution of the problem. The original artificial bee colony algorithm is suitable for solving continuous problems, while vehicle routing problems are discrete problems. At the same time, the artificial bee colony algorithm itself has the defect of strong exploration ability and weak mining ability. Therefore, it is necessary to design and improve the artificial bee colony algorithm to make it suitable for solving vehicle routing problems.

SUMMARY OF THE INVENTION

The present invention specifically considers that the driver's performance in the delivery process is directly affected by the driver's fatigue. This method uses an improved artificial bee colony algorithm to solve this problem, and proposes an urban logistics vehicle route optimization method considering the driver's dynamic efficiency, including the following steps:

S1, determine the relationship between driver fatigue and performance;

S2, establish a vehicle path optimization model;

S3, solve the established above model.

Preferably, the S1, determining the relationship between driver fatigue and performance, specifically includes the following steps:

S11, introduce the fatigue curve model;

S12, analyze the relationship between driver fatigue and performance.

Preferably, the S2, establishing a vehicle path optimization model, specifically includes the following steps:

S21, determine the problem objectives and constraints;

S22, determine the symbolic representation of parameters and variables;

S23, establishing a mathematical model.

Preferably, in said S3, solving the established above-mentioned model specifically includes the following steps:

S31, coding and evaluation of solutions;

S32, initialization stage design;

S33, hiring bee stage design;

S34, follow the bee stage design;

S35, scout bee stage design.

Preferably, in the introduction of the fatigue curve model, the classic fatigue curve model in the field of ergonomics is introduced, and the specific function expression is:

f _t =1-e- ^λt

Among them, f _t is the fatigue level of the driver at time t, and λ is the fatigue index, indicating the speed of fatigue accumulation.

Preferably, in the analysis of the relationship between driver fatigue and performance, the fatigue curve model is converted to determine the driver's delivery speed function under the influence of fatigue, and the formula is:

v _t =v ₀ (1-f _t )=v ₀ e ^-λt

Among them, v _t is the delivery speed of the driver at time t, and v ₀ is the initial speed.

Preferably, it can be known from the driver's delivery speed function that the driver's delivery speed varies nonlinearly, so the relationship function between delivery speed and distance is expressed as:

Among them, d _ij is the distance between customer i and customer j, and s _i and s _j are the time to reach customer i and customer j, respectively;

The time t _ij it takes for the driver to travel from customer point i to customer point j is obtained, the formula is:

Preferably, in determining the problem objectives and constraints, the problem is described as: a distribution center has several vehicles and provides logistics distribution services for multiple customers, where the locations and needs of the customers are known; each customer can only be A vehicle serves and is only served once, and the number of vehicles dispatched cannot exceed the number of vehicles owned by the distribution center; each vehicle has a capacity limit, so the sum of customer demands on each distribution route cannot exceed the maximum capacity of the vehicle; The vehicles all start from the distribution center and must return to the distribution center after serving all the customers on the route; it is assumed that the driver's fatigue does not exist at the initial moment, but as time goes on, the fatigue continues to accumulate. Affected, the driver's performance during the delivery process has changed; in addition, the driver's continuous delivery time is limited to a preset time.

Preferably, the symbolic representation of the determined parameters and variables includes the following parameters and parameter descriptions:

G city logistics distribution network, G = (V, A);

V node set, V={0,1,2,...,n}, where 0 is the distribution center, and 1-n is the customer point;

A arc set, consisting of routes between nodes, A={(i,j)|i,j∈V,i≠j};

A set of K vehicles, K={1,2,...,m}, where vehicles correspond to drivers one-to-one;

d _ij distance between node i and node j;

QThe maximum capacity of the vehicle;

q _i is the demand of customer i;

t _ij vehicle travel time from node i to node j;

The time when the vehicle s _i arrives at node i;

The longest continuous delivery time allowed by the driver;

x _ijk 0-1 variable, 1 when vehicle k travels from node i to node j, and 0 otherwise;

y _ik 0-1 variable, 1 when vehicle k asks node i, and 0 otherwise.

Preferably, the establishing a mathematical model includes establishing a mathematical model for vehicle path optimization according to the problem objective, constraints and symbolic representation:

Equation (1) is the objective function, which means minimizing the total delivery time; Equation (2) indicates that each customer can only be served by one vehicle; Equation (3) indicates that if vehicle k serves customer j, it must visit the node j; Equation (4) indicates that if vehicle k provides service for customer i, it must leave customer i after the service; Equation (5) indicates that vehicles all start from the distribution center and must return to the route after serving all customers on the route. distribution center, and each vehicle only travels along one distribution route; Equation (6) indicates that the sum of all customer demands on each distribution path cannot exceed the maximum capacity of the vehicle; Equation (7) indicates that the number of vehicles used cannot exceed the number of vehicles owned by the distribution center Equation (8) represents the time relationship between vehicles arriving two customers in succession; Equation (9) indicates that the delivery time of each vehicle does not exceed the allowable continuous delivery time; Equations (10) and (11) are the decision variables to take value constraints.

The beneficial effects of the present invention at least include:

1) The present invention considers the dynamic change of the driver's delivery performance caused by fatigue from the perspective of humanistic care when optimizing the vehicle route, and introduces a delivery speed function based on the degree of fatigue to measure this important change. The vehicle path optimization model is constructed with the goal of the shortest time, and the improved artificial bee colony algorithm is used to solve the optimization model;

2) The present invention not only enables the logistics enterprise to plan the driver's distribution path more reasonably in the case of limited resources, helps the enterprise to improve the distribution efficiency and save the distribution cost, but also makes the path planning process take into account the interests of the driver and alleviate the driving problem. It can reduce the fatigue accumulation of the driver and improve the working experience of the driver. These two aspects are of great significance for improving the competitiveness of enterprises.

Description of drawings

FIG. 1 is a flow chart of steps of a method for optimizing the route of an urban logistics vehicle considering the dynamic efficiency of a driver according to an embodiment of the present invention;

Fig. 2 is the S3 step flow chart of the urban logistics vehicle route optimization method considering the driver's dynamic efficiency according to the embodiment of the present invention;

Fig. 3 is the coding schematic diagram of the urban logistics vehicle route optimization method considering the driver's dynamic efficiency according to the embodiment of the present invention;

4 is a schematic diagram of an operator of an urban logistics vehicle route optimization method considering driver dynamic efficiency according to an embodiment of the present invention;

5 is a comparison diagram of an urban logistics vehicle route optimization method considering driver dynamic efficiency according to an embodiment of the present invention and the prior art;

FIG. 6 is a vehicle route planning diagram of an urban logistics vehicle route optimization method considering driver dynamic efficiency according to an embodiment of the present invention.

Detailed ways

In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, but not to limit the present invention.

On the contrary, the present invention covers any alternatives, modifications, equivalents and arrangements within the spirit and scope of the present invention as defined by the appended claims. Further, in order to give the public a better understanding of the present invention, some specific details are described in detail in the following detailed description of the present invention. The present invention can be fully understood by those skilled in the art without the description of these detailed parts.

Referring to Fig. 1, it is a flow chart of the steps of the method, including the following steps:

S1, determine the relationship between driver fatigue and performance;

S11, introduce the fatigue curve model;

In the present invention, the driver's performance in the delivery process is directly affected by the important factor of driver fatigue. Fatigue is a physiological phenomenon, and its current research mainly focuses on the field of ergonomics. In the process of urban logistics and distribution, the driver maintains the same driving state for a long time, which will lead to the appearance of fatigue. At the same time, fatigue builds up over time. According to relevant research, fatigue is generally described as a function of time, that is, the degree of fatigue continues to increase over time. Here, the present invention introduces the classical fatigue curve model in the field of ergonomics, and the specific function expression is:

f _t =1-e- ^λt

Among them, f _t is the fatigue degree of the driver at time t, and λ is the fatigue index, indicating the speed of fatigue accumulation.

S12, analyze the relationship between driver fatigue and performance.

When the dispatcher is planning the vehicle route, it is often the default that the delivery efficiency of the driver is constant. However, according to relevant studies, the fatigue generated during delivery will reduce the driver's work ability, which has a negative impact on the driver's work performance, and at the same time shows a decreasing trend in terms of driving speed. In the present invention, the change of the driver's delivery speed can be used as an index to measure the driver's dynamic performance. Therefore, with the continuous accumulation of driver fatigue, the dynamic performance of the driver can be characterized as a decreasing trend in the delivery speed. The invention converts the fatigue curve model to determine the driver's delivery speed function under the influence of fatigue, and the formula is as follows:

v _t =v ₀ (1-v _t )=v ₀ e ^-λt

where v _t is the delivery speed of the driver at time t, and v ₀ is the initial speed.

In general, changes in driver speed ultimately impact delivery times. It can be seen from the above delivery speed function that the driver's delivery speed varies nonlinearly, so the relationship function between delivery speed and distance can be expressed as:

Where d _ij is the distance between customer i and customer j, and s _i and s _j are the time to reach customer i and customer j, respectively.

Finally, the present invention obtains the time t _ij spent by the driver from customer point i to customer point j, and the formula is as follows:

Through the introduction of the driver fatigue curve model, the analysis and derivation of the driver's delivery speed and the driving time formula, the dynamic performance of the driver in the delivery process can be expressed by a clear functional relationship, which is conducive to the impact of complex driver fatigue. The dynamic performance of the following is quantified.

S2, establish a vehicle path optimization model;

S21, determine the problem objectives and constraints;

The optimization goal of the present invention is the shortest total distribution time. The specific problem is described as follows: a distribution center has several vehicles and provides logistics distribution services for multiple customers, where the location and demand of the customers are known. Each customer can only be serviced by one vehicle and only once, and the number of vehicles dispatched cannot exceed the number of vehicles owned by the distribution center. Each vehicle has a capacity limit, so the sum of customer demand on each delivery route cannot exceed the maximum capacity of the vehicle. Each vehicle departs from the distribution center and must return to the distribution center after serving all customers in its path. It is assumed that driver fatigue does not exist at the initial moment, but it accumulates over time. The driver's performance during delivery has changed due to fatigue during delivery. In addition, combined with relevant regulations, the continuous delivery time of the driver should be limited to a reasonable time.

S22, determine the symbolic representation of parameters and variables;

In order to facilitate the establishment of the subsequent mathematical model, it is necessary to explain the parameters and variables involved in the model.

G city logistics distribution network, G = (V, A);

V node set, V={0,1,2,...,n}, where 0 is the distribution center, and 1-n is the customer point;

A arc set, consisting of routes between nodes, A={(i,j)|i,j∈V,i≠j};

A set of K vehicles, K={1,2,...,m}, where vehicles correspond to drivers one-to-one;

d _ij distance between node i and node j;

QThe maximum capacity of the vehicle;

q _i is the demand of customer i;

t _ij vehicle travel time from node i to node j;

The time when the vehicle s _i arrives at node i;

The longest continuous delivery time allowed by the driver;

x _ijk 0-1 variable, 1 when vehicle k travels from node i to node j, 0 otherwise;

y _ik 0-1 variable, it is 1 when vehicle k asks node i, otherwise it is 0;

S23, establishing a mathematical model.

According to the above problem objectives, constraints and symbolic representations, the present invention finally establishes a mathematical model for vehicle path optimization:

Equation (1) is the objective function, which means minimizing the total delivery time; Equation (2) indicates that each customer can only be served by one vehicle; Equation (3) indicates that if vehicle k serves customer j, it must visit the node j; Equation (4) indicates that if vehicle k provides service for customer i, it must leave customer i after the service; Equation (5) indicates that vehicles all start from the distribution center and must return to the route after serving all customers on the route. distribution center, and each vehicle only travels along one distribution route; Equation (6) indicates that the sum of all customer demands on each distribution path cannot exceed the maximum capacity of the vehicle; Equation (7) indicates that the number of vehicles used cannot exceed the number of vehicles owned by the distribution center Equation (8) represents the time relationship between vehicles arriving two customers in succession; Equation (9) indicates that the delivery time of each vehicle does not exceed the allowable continuous delivery time; Equations (10) and (11) are the decision variables to take value constraints.

S3, solve the established above model.

The artificial bee colony algorithm simulates the behavior of bee colony collecting honey, in which bees collect honey mainly through the following four processes:

(1) At the initial moment, the scout bee randomly searches for the nectar source, and turns into a hired bee after searching for the nectar source.

(2) Hire bees to collect honey and record the relevant information of the nectar source, and at the same time look for a new nectar source with better quality near the nectar source. After the honey collection is over, the hired bees return to the hive to share the information about the nectar source they carry by dancing the swing dance, and recruit the bees to mine the nectar source.

(3) The follower bees who are waiting in the hive select a nectar source to collect nectar according to the information provided by the employed bee, and at the same time look for new and better nectar sources near the nectar source.

(4) The nectar source is exhausted, and the hired bees give up the nectar source and turn into scout bees to find a new nectar source.

Under this honey collection mechanism, bees complete the process of collecting honey throughout the colony through communication and cooperation. According to this process, the basic process of artificial bee colony algorithm can be divided into initialization phase, hiring bee phase, follower bee phase and scout bee phase. The initialization phase is responsible for the setting of various parameters of the algorithm and the generation of the initial solution. The hiring bee phase and the follower bee phase perform search and greedy selection in the solution space, while the scout bee phase replaces many unimproved solutions. Referring to Figure 2, the specific steps of S3 are as follows:

S31, coding and evaluation of solutions;

The nectar source represents the feasible solution of the problem, and the quality of the nectar source represents the fitness value of the solution. In the present invention, the solution of the vehicle routing problem consists of the travel paths of all vehicles. The present invention expresses the solution of the problem by means of natural number coding. Referring to Figure 3, it is assumed that there are 6 customers who need delivery services, and the distribution center has 3 vehicles. In this case, the solution to the problem can be represented by the sequence of Figure 3. The value 0 represents the distribution center, and the non-zero number is the customer number of the vehicle service. A sequence between two values of 0 represents the delivery path of one vehicle, and the delivery path of 3 vehicles constitutes a complete solution to the problem. Therefore, vehicle 1 takes the route 0-1-3-0, vehicle 2 takes the route 0-2-4-0, and vehicle 3 takes the route 0-5-6-0. According to the coding rules, the dimension of each solution is Dim=N+K+1, where N represents the number of customers and K represents the number of vehicles.

Since the search for solutions is random, the algorithm is likely to generate infeasible solutions. In order to maintain the diversity of the population, the present invention allows the existence of infeasible solutions. Therefore, the objective function value cannot be used directly for the evaluation of the solution. The present invention chooses to use the following formula as the evaluation function, that is, on the basis of the original objective function, the penalty cost for exceeding the vehicle capacity is additionally added.

S32, initialization stage design;

Initialize the algorithm parameters such as the number of population CS, the number of employed bees and follower bees SN, the maximum number of iterations MaxCycle, the number of consecutive unimproved trails, and the consecutively unimproved threshold Limit and other algorithm parameters, and generate an initial population. Among them, employed bees and follower bees accounted for half of the population respectively.

In the process of generating the initial population, both the quality and feasibility of the individuals must be considered. The quality of the individual affects the convergence speed of the algorithm, and each individual must satisfy the constraint that the sum of customer demands is not greater than the vehicle capacity. The present invention adopts a simple insertion heuristic method to generate the initial population, and the specific steps are as follows:

(1) Perform multiple random exchanges on a set of sequence 1 containing all client points to obtain sequence 2;

(2) Add a 0 value representing the distribution center at the initial position of an empty sequence 3;

(3) Judging the demand of the customer points in sequence 2 that are not added to sequence 3. If the sum of the demand on the current sub-path after adding the customer point is not greater than the vehicle capacity, the customer point is added successfully, otherwise the customer is skipped. point;

(4) Judging and adding customer points to sequence 3 in turn, if there is no customer point to be added, the customer point addition on the current sub-path ends, and finally adds a value of 0;

(5) Repeat steps (3) and (4) to generate all sub-paths.

After the initial population is generated, the fitness value fit of each individual needs to be calculated. The calculation formula is as follows:

In the problem model, the goal is to minimize delivery time. Therefore, the better the individual, the smaller the objective function value. According to the fitness value calculation formula, the smaller the objective function value is, the larger the individual fitness value is, which is consistent with the basic principle of the artificial bee colony algorithm.

S33, hiring bee stage design;

The hiring bee stage draws on the idea of breadth-first search and focuses on extensive search in the search space to expand the search space of the solution. In order to highlight the role played by this search strategy, the present invention divides the search operators into two categories. The first type of operator is the operator that makes the difference between the individual before and after the update larger, and the second type is the operator that makes the difference between the individual before and after the update smaller. son. For each individual in the population, the hired bee stage uses the first-class operator to perform a neighborhood search. The first type of operator mainly involves the change of customer points between sub-paths or the cross change between different individuals. These operators will make the difference between individuals before and after iteration larger, so the stage of hiring bees can achieve the purpose of extensive search in the search space. There are four operators in the first type of operator, which are exchange, insertion, reversal and crossover. The specific operations are shown in Figure 4, where (a) is exchange, (b) is insertion, (c) is reversal, (d) ) is a cross. The exchange operator randomly selects one to three client points in different sub-paths to exchange. The insertion operator randomly selects one to three customer points in one sub-path and inserts them into another sub-path. The reversal operator randomly selects a range of client points and then reverses their order. The crossover operator draws on the selection and crossover operation of the genetic algorithm. During the iteration, it first retains a sub-path whose total demand is closest to the vehicle capacity in the individual, and then randomly selects another individual in the population to remove the customer points that are duplicated with the reserved part. , and finally merge the two to generate a new individual.

S34, follow the bee stage design;

The following bee stage draws on the idea of depth-first search, focusing on in-depth development in the search space to speed up the convergence speed of the algorithm. The follower bee phase first selects an individual, and then performs neighborhood search with the second-class operator. The second type of operator mainly changes the customer points in the sub-path. These operators are similar to the three types of operators of exchange, insertion and reversal in the hiring bee phase. Only one customer point is selected during the exchange and insertion operations, and three types of operators are used. The operator operation is limited to one sub-path of the individual, so the difference before and after individual iteration is small. At the same time, in this stage, after each iteration uses three kinds of operators to search, if the individual is improved, repeat the search in this stage, otherwise go to the next stage. Because of the small degree of individual variation, these operators can be effectively combined with the depth-first search strategy to speed up the convergence of the algorithm. At the same time, since the operator mutation operation is simple, the addition of the depth-first search strategy will not greatly reduce the search speed of the algorithm.

Among them, the individual selection method is the roulette selection method. The basic idea of the roulette selection method is that the probability of each individual in the population being selected is proportional to its fitness value, that is, the greater the fitness value of the individual, the higher the probability of being selected into the next iteration. Therefore, when selecting individuals, the fitness value of each individual in the population should be calculated first, and then the selection probability of each individual should be calculated according to the fitness value. For individual _xi , the calculation formula of its selection probability is as follows:

Then, the cumulative probability of each individual can be obtained through the selection probability of each individual, that is, the sum of the selection probabilities of all individuals before the individual, the formula is as follows:

Finally, a random number m in the interval [0,1] is randomly generated. If q(x _i-1 )<m<q(x _i ), the individual i is selected.

S35, reconnaissance bee stage design;

In the stages of employed bees and follower bees, the fitness value of the new individual and the original individual will be compared. If the fitness value of the new individual generated is smaller, the original individual has not been improved, and the trail value of the number of times that it has not been improved continuously increases by one, otherwise the new individual replaces the original individual and the trail value returns to zero. In the scout bee stage, the algorithm needs to first find an individual with the largest trail value in the population, and then compare the individual's trail value with the threshold limit. If the trail value is greater than the threshold, an individual will be regenerated according to the generation method of the initial solution to replace the individual. This replacement mechanism can not only maintain the diversity of the population, but also improve the ability of the algorithm to jump out of the local optimum.

The technical effect of the present invention is intuitively explained through the following contents, which are mainly divided into two parts: the performance verification of the improved artificial bee colony algorithm and the calculation and analysis of the vehicle path model. These two parts correspond to the first embodiment and the second embodiment respectively.

Embodiment 1;

In order to verify the performance of the improved artificial bee colony algorithm (DBABC) proposed by the present invention, the present invention selects the original artificial bee colony algorithm (OABC) and the improved artificial bee colony algorithm (MABC) proposed in the existing literature for comparative experiments. OABC uses three kinds of operators: exchange, insertion, and reversal, and MABC uses four kinds of operators: exchange, insertion, inversion, and intersection, and the operators used by these two bee colony algorithms have no difference between sub-paths and intra-sub-path changes. In terms of parameter setting, this paper sets the parameters of the three bee colony algorithms as CS=20, Limit=SN*Dim, which can ensure the fairness of the algorithm performance comparison.

First, in this embodiment, E-n51-k5 in the classic CVRP standard test set is selected to compare the optimal value and average value obtained by the three algorithms in 20 runs to evaluate the algorithm performance. See Figure 5 for the convergence speed of the algorithm. It can be seen from the results that the DBABC proposed by the present invention is obviously superior to the other two bee colony algorithms in terms of convergence speed and stability. From the operator's point of view, although the search effect of DBABC may not be as good as that of MABC in the following bee stage, DBABC combines the depth-first search strategy, which significantly speeds up the convergence speed of the algorithm.

Meanwhile, this embodiment selects two sets of CVRP standard test sets (group A and group B) to further evaluate the overall performance of DBABC. As shown in Table 1, the bold values represent the best results among the three algorithms. From the numerical experiment results, the DBABC designed by the present invention has obvious advantages in solving quality and stability. The reason is that the design of the search strategy of the follower bee and the scout bee further balances the search breadth and search depth of the algorithm. It enhances the local search and global search capabilities of the algorithm.

Table 1

Embodiment 2

In this embodiment, a group A data set in the standard test set CVRP is selected to conduct a simulation experiment, and it is determined that the customer points and distribution centers of this group of examples are distributed on the coordinate axis whose abscissa and ordinate are 0-100km. The fatigue index λ is set to 0.05, the continuous delivery time of the driver is set to not exceed 4h, and the initial delivery speed of the driver is 55km/h. In order to study the difference between the optimal delivery schemes before and after considering the influence of driver fatigue, this embodiment uses an improved artificial bee colony algorithm to solve the imbalance of vehicle paths, total delivery time and driver workload in two different experimental scenarios. , where disequilibrium is expressed as the average difference in driver delivery times.

Table 2

The experimental results are shown in Table 2. Compared with the delivery scheme without considering the influence of driver fatigue, the total delivery time increases by about 10.72% on average after considering the influence of fatigue, while the imbalance of driver delivery time decreases by 11.53% on average. In a high-intensity working environment, the impact of fatigue is objective. Since the delivery efficiency of the driver is on the decline during the delivery process, the company is likely to give a delivery time requirement that is difficult for the driver to meet when formulating the delivery plan. For example, in the calculation example A-n38-k5, the driver actually takes 14.12 hours to complete the delivery task because of fatigue. If the effect of fatigue is ignored, the driver is required to complete the delivery within 13.27 hours. At the same time, the imbalance of driver workload in the calculation example A-n38-k5 is reduced from 54.55% to 38.63%, which means that the driver's workload is also more balanced in the optimal path affected by fatigue. It is worth noting that when the fatigue factor is less affected, order allocation is not affected. The degree of fatigue affecting each driver is inconsistent, which leads to a situation where the unevenness of delivery time becomes larger.

In order to better reflect the influence of fatigue on the optimal delivery route of drivers, this part still takes the calculation example A-n38-k5 as an example for detailed description. Referring to Figure 6, comparing the vehicle paths of the two optimal delivery schemes, it can be found that the delivery paths of the two schemes have changed greatly, which is mainly caused by whether the influence of fatigue is considered in the delivery process. On the one hand, the fatigue ceiling constraint limits the continuous delivery time of the driver, which in turn limits the presence of too many customers on the route. On the other hand, if a certain driver's continuous delivery time is too long, the delivery efficiency may be much lower than that of other drivers. In order to minimize the total delivery time, when optimizing the delivery plan, customers will be more inclined to assign customers to drivers with high delivery efficiency.

The above descriptions are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present invention shall be included in the protection of the present invention. within the range.

Claims (10)

1. An urban logistics vehicle route optimization method considering the driver's dynamic efficiency is characterized in that, comprising the following steps: S1, determine the relationship between driver fatigue and performance; S2, establish a vehicle path optimization model; S3, solve the established above model. 2. the urban logistics vehicle route optimization method considering the driver's dynamic efficiency according to claim 1, is characterized in that, described S1, determines the relation between driver's fatigue and performance, specifically comprises the following steps: S11, introduce the fatigue curve model; S12, analyze the relationship between driver fatigue and performance. 3. the urban logistics vehicle route optimization method considering driver dynamic efficiency according to claim 2, is characterized in that, described S2, establishes vehicle route optimization model, specifically comprises the following steps: S21, determine the problem objectives and constraints; S22, determine the symbolic representation of parameters and variables; S23, establishing a mathematical model. 4. the urban logistics vehicle route optimization method considering the driver's dynamic efficiency according to claim 3, is characterized in that, described S3, solves the above-mentioned model established, specifically comprises the following steps: S31, coding and evaluation of solutions; S32, initialization stage design; S33, hiring bee stage design; S34, follow the bee stage design; S35, scout bee stage design. 5. The urban logistics vehicle route optimization method considering the driver's dynamic efficiency according to claim 4, wherein the introduction of the fatigue curve model, the introduction of the classic fatigue curve model in the field of ergonomics, the specific function expression for: f _t =1-e- ^λt Among them, f _t is the fatigue level of the driver at time t, and λ is the fatigue index, indicating the speed of fatigue accumulation. 6. The urban logistics vehicle route optimization method considering the driver's dynamic efficiency according to claim 5, characterized in that, in the analysis of the relationship between driver fatigue and performance, the fatigue curve model is converted to determine the effect of fatigue The driver delivery speed function below, the formula is: v _t =v ₀ (1-f _t )=v ₀ e ^-λt Among them, v _t is the delivery speed of the driver at time t, and v ₀ is the initial speed. 7. The urban logistics vehicle route optimization method considering the driver's dynamic efficiency according to claim 6, characterized in that, from the driver's distribution speed function, it can be known that the driver's distribution speed is nonlinear, so the distribution speed and The relation function of distance is expressed as:

Among them, d _ij is the distance between customer i and customer j, and s _i and s _j are the time to reach customer i and customer j, respectively; The time t _ij it takes for the driver to travel from customer point i to customer point j is obtained, the formula is:

8. The urban logistics vehicle route optimization method considering the driver's dynamic efficiency according to claim 7, wherein, in the determination of the problem objectives and constraints, the problem is described as: a distribution center has several vehicles, which are multiple The customer provides logistics and distribution services, in which the customer's location and needs are known; each customer can only be served by one vehicle and only once, and the number of vehicles dispatched cannot exceed the number of vehicles owned by the distribution center; There is a capacity limit, so the sum of customer demands on each delivery route cannot exceed the maximum capacity of the vehicle; each vehicle departs from the distribution center and must return to the distribution center after serving all customers on the route; it is assumed that the driver at the initial moment Fatigue does not exist, but it builds up over time, and the driver's performance during delivery changes due to being affected by fatigue during delivery; furthermore, the driver's continuous delivery time is limited to a pre-delivery period. within the set time. 9. The urban logistics vehicle route optimization method considering the driver's dynamic efficiency according to claim 8, wherein the symbolic representation of the determined parameters and variables includes the following parameters and parameter descriptions: G city logistics distribution network, G = (V, A); V node set, V={0, 1, 2, ..., n}, where 0 is the distribution center, and 1-n is the customer point; A arc set, consisting of routes between nodes, A={(i, j)|i, j∈V, i≠j}; K vehicle set, K={1, 2, ..., m}, where vehicles correspond to drivers one-to-one; d _ij distance between node i and node j; Q the maximum capacity of the vehicle; q _i is the demand of customer i; t _ij vehicle travel time from node i to node j; The time when the vehicle s _i arrives at node i; T maximum continuous delivery time allowed by the driver; x _ijk 0-1 variable, 1 when vehicle k travels from node i to node j, and 0 otherwise; y _ik 0-1 variable, 1 when vehicle k asks node i, and 0 otherwise. 10. The urban logistics vehicle route optimization method considering the driver's dynamic efficiency according to claim 9, wherein the establishment of the mathematical model comprises, according to the problem objective, the constraint condition and the symbolic representation, establishment of a mathematical model of the vehicle route optimization :

Equation (1) is the objective function, which means minimizing the total delivery time; Equation (2) indicates that each customer can only be served by one vehicle; Equation (3) indicates that if vehicle k serves customer j, it must visit the node j; Equation (4) indicates that if vehicle k provides service for customer i, it must leave customer i after the service is over; Equation (5) indicates that vehicles all start from the distribution center and must return to the route after serving all customers on the route. distribution center, and each vehicle only travels along one distribution route; Equation (6) indicates that the sum of all customer demands on each distribution path cannot exceed the maximum capacity of the vehicle; Equation (7) indicates that the number of vehicles used cannot exceed the number of vehicles owned by the distribution center Equation (8) represents the time relationship of vehicles arriving to two customers in succession; Equation (9) indicates that the delivery time of each vehicle does not exceed the allowable continuous delivery time; Equations (10) and (11) are the decision variables to take value constraints.

Images