Emergency material site selection and path optimization method based on whale optimization algorithm

Abstract

The invention discloses an emergency material site selection and path optimization method based on a whale optimization algorithm, and the method specifically comprises the following steps: firstly, building an emergency material site selection model based on urgency and customer satisfaction, and selecting an optimal distribution point as a distribution center; secondly, establishing an emergency material distribution model based on transportation cost and time conditions, and planning a distribution path of an affected point; and finally, adjusting based on the parameters of the whale algorithm, and storing the path generated by each iteration to obtain a disaster-affected point path planning route and a global path length. According to the method, the emergency material distribution path optimization is realized, the timeliness is ensured, the cost is reduced, and an effective guarantee is provided for the transportation of the emergency materials.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution center site selection and path optimization, and in particular to an emergency material site selection and path optimization method based on a whale optimization algorithm that takes urgency and satisfaction into consideration. Background Art

[0002] As the urban population grows and the degree of urbanization increases, the frequency of emergencies continues to increase. Traditional emergency material site selection and distribution rely on the storage and transportation systems of the government or rescue agencies. After a disaster occurs, the process from demand assessment to material delivery has problems such as poor information transmission, traffic obstruction, and uneven material distribution, resulting in low rescue efficiency. Therefore, studying how to transport materials in a timely manner and improve distribution efficiency has important theoretical value and practical significance.

[0003] For the site selection of emergency materials, the existing methods mainly include mathematical programming methods, methods based on geographic information system (GIS), multi-criteria decision analysis (MCDA) such as analytic hierarchy process (AHP), fuzzy set theory, etc. For the route planning of emergency materials, the existing algorithms can be mainly divided into several categories: shortest path algorithms such as Dijkstra algorithm and A algorithm, network flow algorithms such as Ford-Fulkerson method, heuristic search algorithms such as genetic algorithm, ant colony algorithm, etc., and methods based on artificial intelligence such as deep learning. However, in practical applications, the above algorithms have limitations such as high time complexity, easy to fall into local optimality, and inability to effectively solve specific problems. At present, there is still a lot of room for development in the research of intelligent optimization algorithms, and many researchers hope to find a more ideal algorithm for the optimization of emergency material transportation routes.

[0004] The present invention aims to realize a high-efficiency and high-precision WOA improved algorithm. The main purpose is to improve the efficiency of the algorithm so as to better apply the direction of emergency material transportation, and secondly to improve the accuracy of the whale algorithm and jump out of the local optimum. The biggest feature of the whale optimization algorithm is to use random individuals or optimal individuals to simulate the hunting behavior of humpback whales, and use spirals to simulate the bubble net attack mechanism of humpback whales. However, the original whale algorithm will gradually reduce the fluctuation of changes as the number of iterations increases, which will lead to local optimum. Therefore, the present invention adds nonlinear convergence factors, adaptive search control coefficients, and memory library list strategies to form a new ACSWOA algorithm to prevent the algorithm from falling into local optimum and speed up the convergence of the algorithm. This not only optimizes the emergency material distribution path, but also enhances the practicality of the algorithm. Summary of the invention

[0005] In view of the defects of the prior art or the need for improvement, the present invention provides an emergency material site selection and path optimization method based on the whale optimization algorithm which takes into account urgency and satisfaction. By combining the site selection and distribution model with the whale optimization algorithm, the planning of the emergency material distribution path is realized, thereby improving the distribution efficiency.

[0006] In order to achieve the above object, the technical solution of the present invention is as follows: a method for optimizing the location and path of emergency materials based on a whale optimization algorithm taking into account urgency and satisfaction, the method comprising the following steps:

[0007] Step 1: Establish an emergency material location selection model based on urgency and customer satisfaction, and select the optimal distribution point as the distribution center.

[0008] Step 2: Establish an emergency material distribution model based on transportation costs and time, and plan the distribution route to the disaster-stricken areas.

[0009] Step 3: Adjust the parameters based on the whale algorithm and store the path generated by each iteration.

[0010] Step 4: Obtain the path planning route of the disaster-affected point and the global path length.

[0011] Preferably, the emergency material site selection model described in step 1 is sorted according to urgency and customer satisfaction, and then the optimal distribution point is selected as the distribution center. The specific implementation process is:

[0012] Step 1-1: Establish an urgency model. The TOPSIS method is combined with the entropy weight method to evaluate the urgency of the demand for emergency supplies at the disaster site. Among them, the advantage of the TOPSIS method is that it is simple, flexible and easy to calculate, but it has a certain degree of subjectivity when determining the weight. The entropy weight method is an objective weight analysis method that can reduce the subjectivity of the TOPSIS method. The specific implementation process is as follows:

[0013] Step 1-1-1: Construct the initial decision matrix. Standardize it. The matrix composed of the index values ​​of n material demand points is called the initial decision matrix.

[0014]

[0015] Among them, Xij represents the index value of the jth column index in the i-th row of the material demand point, X is the initial decision matrix, i = 1, 2, .., n, j = 1, 2, .., m.

[0016] Normalize X to avoid errors caused by different dimensions among indicators.

[0017]

[0018] Among them, Cij is the decision matrix, max(X j ) is the largest index value among the indexes in the jth column.

[0019] Step 1-1-2: Determine the indicator weights and establish the weighted decision matrix R.

[0020]

[0021] Among them, R is the weighted decision matrix, which is obtained by multiplying the decision matrix and the weight obtained by the entropy weight method.

[0022] Step 1-1-3: Determine the positive and negative ideal solutions.

[0023] Positive ideal solution R + :

[0024] R + =(R 1 + ,R 2 + ,..,R m + ),R j + =max(R ij )

[0025] Negative ideal solution R - :

[0026] R - =(R 1 - ,R 2 - ,..,R m - ),R j - =min(R ij ) (Formula 4)

[0027] Among them, R + , R + is the set of positive and negative ideal solutions, represents the positive ideal solution of the mth column of the matrix, represents the negative ideal solution of the mth column of the matrix, represents the maximum positive ideal solution in the jth column of the weighted decision matrix, represents the minimum negative ideal solution in the jth column of the weighted decision matrix.

[0028] Step 1-1-4: Solve for the positive and negative ideal solution distances.

[0029]

[0030] Among them, C +, C - Represent the distance of each solution to the positive and negative ideal solutions respectively.

[0031] Step 1-1-5: Calculate the relative proximity of demand points.

[0032]

[0033] Among them, F 1 Indicates the relative closeness of each disaster point to the positive ideal solution, F 1 The larger the value, the more urgent the demand.

[0034] Step 1-2: Establish a customer satisfaction model. This includes hard time windows and soft time windows. When emergency supplies do not arrive within the hard time window but arrive at the disaster site within the soft time window, customer satisfaction is calculated based on the degree of deviation from the hard time window. The customer satisfaction calculation formula is as follows:

[0035]

[0036] Among them, F 2 is customer satisfaction, and the hard time window is (ET g ,LT g ), when the time range arrives, the customer satisfaction is 1; in the soft time window (et g ,lt g ) arrives outside the customer service center, customer satisfaction is 0; T g represents the time when the service vehicle arrives at the disaster site g;

[0037] Step 1-3: Establish an emergency material model based on the urgency and customer satisfaction scores of each disaster site. Normalize the urgency and satisfaction functions, and assign weights to obtain the disaster site with the highest score. This site is selected as the distribution center for rescue. The specific formula is as follows:

[0038]

[0039] Among them, F 1 、F 2 are urgency and satisfaction values, respectively, 1 ,θ 2 is the corresponding weight, and F is the comprehensive score.

[0040] Preferably, the emergency material distribution model described in step 2 should combine transportation cost, vehicle travel time cost, carbon value cost and government subsidy cost to obtain short delivery time and low transportation cost as constraints. The specific implementation process is:

[0041] Step 2-1: Establish a transportation cost model based on risk management. The transportation cost is reflected in the direct transportation cost related to the transportation quantity, time, and distance during the transportation process, as well as the indirect transportation cost related to risk management. As the transportation time and distance increase, the risks faced by the goods during transportation are higher because more uncertain factors may be encountered, such as weather changes and road conditions. Similarly, the greater the weight of the goods, the greater the potential losses of potential accidents (such as the risk of goods loss) during transportation, and the corresponding risk management cost will also increase. Therefore, establish a transportation cost model based on risk management:

[0042]

[0043] Among them, C 1 represents the transportation cost, d iq represents the distance from the i-th point to the q-th point; U represents the unit distance cost from the i-th point to the q-th point; G represents the quantity, quality, or volume of the transported goods; R r is the risk area, R low is the low-risk demarcation point, R mid is the medium-risk demarcation point; C low is the fixed cost in the low-risk interval; and b mid is the slope and intercept in the medium-risk interval; m high and b high are the slope and intercept in the high-risk interval;

[0044] Step 2-2: Establish a vehicle driving time model. The time cost is used to describe the penalty caused by the vehicle allocation time exceeding the expected time due to road congestion. The shortest time that the vehicle can travel under unobstructed road conditions is used as the expected time. Based on this, the research established a vehicle driving time model function:

[0045]

[0046] Among them, C 2 represents the time cost, d iq is the distance from i to the q-th point; v iq is the speed; α is the cost coefficient of time delay; in order to quantitatively describe the driving situation of vehicles under congested road conditions, combined with the "China Road Capacity Guide" and the low characteristics of road traffic, R is defined as the "road traffic flow evaluation coefficient", and the road traffic conditions are divided into five levels: 0.00 < R < 0.30 indicates unobstructed state; 0.30 < R < 0.60 indicates slightly congested state; 0.60 < R < 0.75 indicates somewhat congested state; 0.75 < R < 0.90 indicates relatively congested state; 0.90 < R < 1.00 indicates congested state; R >= 1.00 indicates severely congested state.

[0047] Step 2-3: Establish a carbon cost model. Carbon emission costs are closely related to multiple variables in the transportation process, including the vehicle load, transportation distance, and fuel consumption. In today's global environment of jointly addressing climate change and reducing greenhouse gas emissions, this model aims to achieve a win-win goal of economic benefits and environmental protection.

[0048]

[0049] Among them, C 3 represents the carbon cost, P * is the amount of fuel consumed per unit mileage; P 0 Q is the amount of fuel consumed per unit mileage when the vehicle is not loaded with cargo; * is the maximum load of the vehicle; Q iq is the current vehicle load.

[0050] Step 2-4: Establish a government subsidy model. During the distribution of emergency supplies, the government will provide corresponding subsidies based on the urgency of the situation. By introducing the midpoint offset parameter μ into the formula, the specific value of the urgency of supplies can be adjusted when the subsidy amount begins to increase significantly.

[0051]

[0052] Among them, S(P) represents the government subsidy cost, and C 4 To express, P is the urgency of materials, 0≤P≤1; k is the subsidy amplification factor, which is used to adjust the overall level of subsidy amount; β is the slope parameter, which controls the rate at which the subsidy amount changes with P; the larger the β, the faster the change; the smaller the β, the smoother the change; μ is the midpoint offset parameter, which is used to adjust the position of the center point of the logistic stick function. When μ=0.5, the function is symmetrical about P=0.5; when μ deviates from 0.5, the function will shift to the left or right; b is the base value or minimum value of the subsidy.

[0053] Step 2-5: A total cost model is obtained based on transportation costs, vehicle travel time costs, carbon value costs and government subsidies. The calculation formula is as follows:

[0054] C(cost)=C 1 +C 2 +C 3 -C 4 (Formula 5)

[0055] Among them, C(cost) is the total cost; C 1 is the transportation cost; C 2 is the vehicle travel time cost; C 3 is the carbon cost; C 4 The government subsidizes the cost.

[0056] Preferably, a new whale optimization algorithm (ACSWOA) is proposed in step 3. For the original whale (WOA) algorithm, a nonlinear convergence factor is added, an adaptive search control coefficient is introduced, and a memory library list strategy is introduced. The specific process is as follows:

[0057] Step 3-1: Improve the original WOA algorithm parameter a, add a nonlinear convergence factor, slow down the algorithm from falling into the local optimum as the number of iterations increases; at the same time, as the number of iterations gradually increases, the convergence speed is effectively accelerated. The changed algorithm parameter a is as follows:

[0058] a=(a start -a end )+(1-t / t max ) / (1-μ 1 *t / t max )(Formula 1)

[0059] Among them, a start and a end are the initial value and final value of a respectively; μ 1 is the adjustment coefficient, take μ 1 =25, M is the maximum number of iterations, and t is the current number of iterations.

[0060] Step 3-2: Introduce an adaptive search control coefficient to dynamically adjust the size of C according to the number of iterations. In the early stages of the algorithm, the C value is small, and the movement of the whale increases, which helps improve the global search capability. As the algorithm iterates, the movement of the whale in the search space decreases, which helps improve the local search capability. The expression is as follows:

[0061]

[0062] Among them, C is the adaptive search control coefficient, r is the coefficient, which can control the decay rate of C, t max is the maximum number of iterations, and t is the current number of iterations.

[0063] Step 3-3: Introduce the memory list strategy. Add a memory list to record the optimal solution encountered in each iteration. After reaching the maximum number of iterations, compare the optimal solution obtained in the last iteration with the solution in the memory, and select the better solution as the final result. To avoid excessive memory capacity, limit the length of the memory list, and remove the worst solution when the maximum capacity is reached. The expression is as follows:

[0064] S=remove(min(S)) (Formula 3)

[0065] Among them, each element stored in the S matrix is ​​the current optimal solution encountered during the iteration process. When the memory library reaches the maximum value, the minimum value in the library is removed by remove.

[0066] Preferably, the path planning route and the global path length of the disaster point are obtained in step 4. The specific process is as follows:

[0067] Step 4-1: Set ACSWOA algorithm parameters: population size N, maximum number of iterations t max , the current iteration number t and the spiral shape constant b, and randomly initialize the initial position Xi (i=1, 2, ..., n) of the whale group;

[0068] Step 4-2: Calculate the fitness of each whale individual according to the following formula, sort the fitness, and find the whale individual corresponding to the minimum fitness, that is, the best whale individual X * And save it to the memory library list. When the library reaches the maximum capacity, remove the worst solution;

[0069]

[0070] S = remove(min(S))

[0071] (Formula 2)

[0072] Step 4-3: Calculate the parameter a that introduces the nonlinear convergence factor and update the coefficient vector A;

[0073] a=(a start -a end )+(1-t / t max ) / (1-μ*t / t max )

[0074] (Formula 3)

[0075] Step 4-4: Introduce adaptive search control coefficients and update coefficient vector C;

[0076]

[0077] Step 4-5: Using α j =min(X i,j ),β j =max(X i,j ) Calculate the dynamic search boundary at this time;

[0078] Step 4-6: Calculate the value of each whale individual, compare the solution of the reverse population with the solution in the current group, select s whale individuals with high fitness as the next generation group, and locate the whale individual with the minimum fitness as the prey position;

[0079] Step 4-7: If t≤t max , then update the parameters a, A, C, l and p;

[0080] Step 4-8: When p < 0.5, if |A| < 1, update the spatial position of the current whale individual according to formula (5); when |A| > = 1, randomly select the whale individual position X rand , and update the spatial position of the current whale individual according to formula (6);

[0081] D=|C·X * (t)-X(t),X(t+1)=X(t)-A·D (Formula 5)

[0082] D=|C·X rand (t)-X|,X(t+1)=X rand -A·D (Formula 6)

[0083] Among them, X*(t) is the position of the best solution currently obtained, X(t) is the current whale position, X(t+1) is the updated whale position, and D is the distance between the individual whale and the current optimal solution.

[0084] Step 4-9: When p>=0.5, update the spatial position of the current individual whale according to formula (7);

[0085]

[0086] Among them, D' is the distance between the current search individual and the current optimal solution, l is a random number with a uniform distribution in the range [-1,1], and b is the spiral shape parameter.

[0087] Step 4-10: Determine the new position according to the constraints. If the new position is feasible, the position of the individual whale will be updated. If not, the position of the individual whale will not change.

[0088] Step 4-11: Determine the termination condition of the algorithm: If t = t max , then go to step 4-10; otherwise, repeat steps 4-7 to 4-11;

[0089] Step 4-12: Output the spatial position X of the optimal whale individual * and its fitness, that is, output the emergency material distribution path and calculate the path length.

[0090] Preferably, an emergency material site selection and path optimization system considering urgency and satisfaction based on the whale optimization algorithm comprises:

[0091] The site selection and sorting module is used to sort the needs and satisfaction of different disaster-affected sites, and the disaster-affected site with the highest score is selected as the distribution center;

[0092] The material distribution module is used to plan the material distribution route and obtain the low-cost and fast distribution route based on the transportation cost, vehicle travel time cost, carbon value cost and government subsidy;

[0093] The algorithm optimization module is used for parameter adjustment and data storage based on the whale optimization algorithm, and applies the improved algorithm to path planning to obtain the path planning route and the global path length.

[0094] In general, the above technical solutions conceived by the present invention can achieve the following beneficial effects compared with the prior art:

[0095] 1. Through this emergency material site selection and path optimization method based on the whale optimization algorithm that considers urgency and satisfaction, the emergency material site selection model combines material urgency and user satisfaction to select the optimal disaster-stricken point as the distribution center, so that materials can be delivered to the disaster-stricken area quickly and efficiently, reducing the losses of disaster victims. It can take into account the differences in material shortages at different disaster-stricken points to avoid oversupply and undersupply. At the same time, taking into account the satisfaction of residents in the disaster-stricken areas with emergency rescue, it will help the Emergency Management Bureau to better serve the people.

[0096] 2. Through the emergency material site selection and path optimization method based on the whale optimization algorithm that considers urgency and satisfaction, the emergency material distribution model combines transportation costs, vehicle travel time costs, carbon value costs and government subsidy costs, which can achieve timely delivery and low cost, thereby improving distribution efficiency.

[0097] 3. Through this method of emergency material location and path optimization based on the whale optimization algorithm that considers urgency and satisfaction, in view of the problem that the original whale algorithm is prone to fall into the local optimum, it is proposed to introduce nonlinear convergence factors and adaptive search control coefficients to balance the relationship between exploration and development and avoid falling into the local optimum too early. And the memory list strategy is used to avoid repeated calculations and speed up the convergence process. By applying the ACSWOA algorithm to the emergency material transportation model and generating the global optimal path, the total cost and time of the transportation process can be controlled, and low carbon can be better guaranteed, which has a certain role in promoting the development of green logistics in my country. BRIEF DESCRIPTION OF THE DRAWINGS

[0098] Figure 1 It is a flow chart of a method for optimizing the location and path of emergency materials based on a whale optimization algorithm taking into account urgency and satisfaction, provided in an embodiment of the present invention;

[0099] Figure 2 It is a module division diagram of an emergency material site selection and path optimization system based on a whale optimization algorithm taking into account urgency and satisfaction, provided in an embodiment of the present invention.

[0100] Figure 3 This is a flow chart of delivery route planning based on the whale optimization algorithm. Specific implementation methods

[0102] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0103] Example 1: Figure 1 and Figure 3 The flowchart of a method for optimizing the location and path of emergency materials based on the whale optimization algorithm considering urgency and satisfaction is shown in an example of the present invention. The method includes the following steps:

[0104] Step 1: Establish an emergency material location selection model based on urgency and customer satisfaction, and select the optimal distribution point as the distribution center.

[0105] Step 1-1: Establish an urgency model. The TOPSIS method is combined with the entropy weight method to evaluate the urgency of the demand for emergency supplies at the disaster site. Among them, the advantage of the TOPSIS method is that it is simple, flexible and easy to calculate, but it has a certain degree of subjectivity when determining the weight. The entropy weight method is an objective weight analysis method that can reduce the subjectivity of the TOPSIS method. The specific implementation process is as follows:

[0106] Step 1-1-1: Construct the initial decision matrix. Standardize it. The matrix composed of the index values ​​of n material demand points is called the initial decision matrix.

[0107]

[0108] Among them, Xij represents the index value of the jth column index in the i-th row of the material demand point, X is the initial decision matrix, i = 1, 2, .., n, j = 1, 2, .., m.

[0109] Normalize X to avoid errors caused by different dimensions among indicators.

[0110]

[0111] Among them, C ij is the decision matrix, max(X j ) is the largest index value among the indexes in the jth column.

[0112] Step 1-1-2: Determine the indicator weights and establish the weighted decision matrix R.

[0113]

[0114] Among them, R is the weighted decision matrix, which is obtained by multiplying the decision matrix and the weight obtained by the entropy weight method.

[0115] Step 1-1-3: Determine the positive and negative ideal solutions.

[0116] Positive ideal solution R + :

[0117] R + =(R 1 + ,R 2 + ,..,R m + ),R j + =max(R ij )

[0118] Negative ideal solution R - :

[0119] R - =(R 1 - ,R 2 - ,..,R m - ),R j - =min(R ij ) (Formula 4)

[0120] Among them, R + , R + is the set of positive and negative ideal solutions, represents the positive ideal solution of the mth column of the matrix, represents the negative ideal solution of the mth column of the matrix, represents the maximum positive ideal solution in the jth column of the weighted decision matrix, represents the minimum negative ideal solution in the jth column of the weighted decision matrix.

[0121] Step 1-1-4: Solve for the positive and negative ideal solution distances.

[0122]

[0123] Among them, C + , C - Represent the distance of each solution to the positive and negative ideal solutions respectively.

[0124] Step 1-1-5: Calculate the relative proximity of demand points.

[0125]

[0126] Among them, F 1 Indicates the relative closeness of each disaster point to the positive ideal solution, F 1 The larger the value, the more urgent the demand.

[0127] Step 1-2: Establish a customer satisfaction model. This includes hard time windows and soft time windows. When emergency supplies do not arrive within the hard time window but arrive at the disaster site within the soft time window, customer satisfaction is calculated based on the degree of deviation from the hard time window. The customer satisfaction calculation formula is as follows:

[0128]

[0129] Among them, F 2 is customer satisfaction, and the hard time window is (ET g ,LT g ), when the time range arrives, the customer satisfaction is 1; in the soft time window (et g ,lt g ) arrives outside the customer service center, customer satisfaction is 0; T g represents the time when the service vehicle arrives at the disaster site g;

[0130] Step 1-3: Establish an emergency material model based on the urgency and customer satisfaction scores of each disaster site. Normalize the urgency and satisfaction functions, and assign weights to obtain the disaster site with the highest score. This site is selected as the distribution center for rescue. The specific formula is as follows:

[0131]

[0132] Among them, F 1 、F 2 are urgency and satisfaction values, respectively, 1 ,θ 2 is the corresponding weight, and F is the comprehensive score.

[0133] Step 2: Establish an emergency material distribution model based on transportation costs and time, and plan the distribution route to the disaster-stricken areas.

[0134] The emergency material distribution model should combine transportation cost, vehicle travel time cost, carbon value cost and government subsidy cost to obtain short delivery time and low transportation cost as constraints. The specific steps are as follows:

[0135] Step 2-1: Establish a transportation cost model based on risk management. Transportation costs are reflected in the direct transportation costs related to the quantity, time, and distance during transportation, as well as the indirect transportation costs related to risk management. As the transportation time and distance increase, the risks faced by goods during transportation are higher because more uncertain factors may be encountered, such as weather changes and road conditions. Similarly, the greater the weight of the goods, the greater the potential losses of potential accidents (such as the risk of goods loss) during transportation, and the corresponding risk management costs will also increase. Therefore, establish a transportation cost model based on risk management:

[0136]

[0137] Among them, C 1 represents the transportation cost, d iq represents the distance from the i-th point to the q-th point; U represents the unit distance cost from the i-th point to the q-th point; G represents the quantity, quality, or volume of the transported goods; R r is the risk area, R low is the low-risk demarcation point, R mid is the medium-risk demarcation point; C low is the fixed cost in the low-risk interval; and b mid is the slope and intercept in the medium-risk interval; m high and b high are the slope and intercept in the high-risk interval;

[0138] Step 2-2: Establish a vehicle driving time model. The time cost is used to describe the penalty caused by the vehicle allocation time exceeding the expected time due to road congestion. The shortest time that the vehicle can travel under unobstructed road conditions is used as the expected time. Based on this, the research established a vehicle driving time model function:

[0139]

[0140] Among them, C 2 represents the time cost, d iq is the distance from i to the q-th point; v iq is the speed; α is the cost coefficient of time delay; in order to quantitatively describe the driving situation of vehicles under congested road conditions, combined with the "China Road Capacity Guide" and the low characteristics of road traffic, R is defined as the "road traffic flow evaluation coefficient", and the road traffic conditions are divided into five levels: 0.00 < R < 0.30 indicates unobstructed state; 0.30 < R < 0.60 indicates slightly congested state; 0.60 < R < 0.75 indicates somewhat congested state; 0.75 < R < 0.90 indicates relatively congested state; 0.90 < R < 1.00 indicates congested state; R ≥ 1.00 indicates severely congested state.

[0141] Step 2-3: Establish a carbon cost model. Carbon emission costs are closely related to multiple variables in the transportation process, including the vehicle load, transportation distance, and fuel consumption. In today's global environment of jointly addressing climate change and reducing greenhouse gas emissions, this model aims to achieve a win-win goal of economic benefits and environmental protection.

[0142]

[0143] Among them, C 3 represents the carbon cost, P * is the amount of fuel consumed per unit mileage; P 0 Q is the amount of fuel consumed per unit mileage when the vehicle is not loaded with cargo; * is the maximum load of the vehicle; Q iq is the current vehicle load.

[0144] Step 2-4: Establish a government subsidy model. During the distribution of emergency supplies, the government will provide corresponding subsidies based on the urgency of the situation. By introducing the midpoint offset parameter μ into the formula, the specific value of the urgency of supplies can be adjusted when the subsidy amount begins to increase significantly.

[0145]

[0146] Among them, S(P) represents the government subsidy cost, and C 4 To express, P is the urgency of materials, 0≤P≤1; k is the subsidy amplification factor, which is used to adjust the overall level of subsidy amount; β is the slope parameter, which controls the rate at which the subsidy amount changes with P; the larger the β, the faster the change; the smaller the β, the smoother the change; μ is the midpoint offset parameter, which is used to adjust the position of the center point of the logistic stick function. When μ=0.5, the function is symmetrical about P=0.5; when μ deviates from 0.5, the function will shift to the left or right; b is the base value or minimum value of the subsidy.

[0147] Step 2-5: A total cost model is obtained based on transportation costs, vehicle travel time costs, carbon value costs and government subsidies. The calculation formula is as follows:

[0148] C(cost)=C 1 +C 2 +C 3 -C 4 (Formula 5)

[0149] Among them, C(cost) is the total cost; C 1 is the transportation cost; C 2 is the vehicle travel time cost; C 3 is the carbon cost; C 4 The government subsidizes the cost.

[0150] Step 3: Adjust the parameters based on the whale algorithm and store the path generated by each iteration.

[0151] For the original Whale (WOA) algorithm, a nonlinear convergence factor is added, an adaptive search control coefficient is introduced, and a memory library list strategy is introduced. The specific steps are as follows:

[0152] Step 3-1: Improve the original WOA algorithm parameter a, add a nonlinear convergence factor, slow down the algorithm from falling into the local optimum as the number of iterations increases; at the same time, as the number of iterations gradually increases, the convergence speed is effectively accelerated. The changed algorithm parameter a is as follows: a = (a start -a end )+(1-t / t max ) / (1-μ 1 *t / t max )(Formula 1)

[0153] Among them, a start and a end are the initial value and final value of a respectively; μ 1 is the adjustment coefficient, take μ 1 =25, M is the maximum number of iterations, and t is the current number of iterations.

[0154] Step 3-2: Introduce an adaptive search control coefficient to dynamically adjust the size of C according to the number of iterations. In the early stages of the algorithm, the C value is small, and the movement of the whale increases, which helps improve the global search capability. As the algorithm iterates, the movement of the whale in the search space decreases, which helps improve the local search capability. The expression is as follows:

[0155]

[0156] Among them, C is the adaptive search control coefficient, r is the coefficient, which can control the decay rate of C, t max is the maximum number of iterations, and t is the current number of iterations.

[0157] Step 3-3: Introduce the memory list strategy. Add a memory list to record the optimal solution encountered in each iteration. After reaching the maximum number of iterations, compare the optimal solution obtained in the last iteration with the solution in the memory, and select the better solution as the final result. To avoid excessive memory capacity, limit the length of the memory list, and remove the worst solution when the maximum capacity is reached. The expression is as follows:

[0158] S=remove(min(S)) (Formula 3)

[0159] Among them, each element stored in the S matrix is ​​the current optimal solution encountered during the iteration process. When the memory library reaches the maximum value, the minimum value in the library is removed by remove.

[0160] Step 4: Obtain the path planning route of the disaster-affected point and the global path length.

[0161] Step 4-1: Set ACSWOA algorithm parameters: population size N, maximum number of iterations t max , the current iteration number t and the spiral shape constant b, and randomly initialize the initial position Xi (i=1, 2, ..., n) of the whale group;

[0162] Step 4-2: Calculate the fitness of each whale individual according to the following formula, sort the fitness, and find the whale individual corresponding to the minimum fitness, that is, the best whale individual X * And save it to the memory library list. When the library reaches the maximum capacity, remove the worst solution;

[0163]

[0164] S = remove(min(S))

[0165] (Formula 2)

[0166] Step 4-3: Calculate the parameter a that introduces the nonlinear convergence factor and update the coefficient vector A;

[0167] a=(a start -a end )+(1-t / t max ) / (1-μ*t / t max )

[0168] (Formula 3)

[0169] Step 4-4: Introduce adaptive search control coefficients and update coefficient vector C;

[0170]

[0171] Step 4-5: Using α j =min(X i,j ),β j =max(X i,j ) Calculate the dynamic search boundary at this time;

[0172] Step 4-6: Calculate the value of each whale individual, compare the solution of the reverse population with the solution in the current group, select s whale individuals with high fitness as the next generation group, and locate the whale individual with the minimum fitness as the prey position;

[0173] Step 4-7: If t≤t max, then update the parameters a, A, C, l and p;

[0174] Step 4-8: When p < 0.5, if |A| < 1, update the spatial position of the current whale individual according to formula (5); when |A| > = 1, randomly select the whale individual position X rand , and update the spatial position of the current whale individual according to formula (6);

[0175] D=|C·X * (t)-X(t),X(t+1)=X(t)-A·D (Formula 5)

[0176] D=|C·X rand (t)-X|,X(t+1)=X rand -A·D (Formula 6)

[0177] Among them, X*(t) is the position of the best solution currently obtained, X(t) is the current whale position, X(t+1) is the updated whale position, and D is the distance between the individual whale and the current optimal solution.

[0178] Step 4-9: When p>=0.5, update the spatial position of the current individual whale according to formula (7);

[0179]

[0180] Among them, D' is the distance between the current search individual and the current optimal solution, l is a random number with a uniform distribution in the range [-1,1], and b is the spiral shape parameter.

[0181] Step 4-10: Determine the new position according to the constraints. If the new position is feasible, the position of the individual whale will be updated. If not, the position of the individual whale will not change.

[0182] Step 4-11: Determine the termination condition of the algorithm: If t = t max , then go to step 4-10; otherwise, repeat steps 4-7 to 4-11;

[0183] Step 4-12: Output the spatial position X of the optimal whale individual * and its fitness, that is, output the emergency material distribution path and calculate the path length.

[0184] Example 2: Figure 2 As shown, an embodiment of the present invention provides an emergency material site selection and path optimization system based on a whale optimization algorithm that considers urgency and satisfaction. The system includes the following modules:

[0185] The site selection and sorting module is used to sort the needs and satisfaction of different disaster-affected sites, and the disaster-affected site with the highest score is selected as the distribution center;

[0186] The material distribution module is used to plan the material distribution route and obtain the low-cost and fast distribution route based on the transportation cost, vehicle travel time cost, carbon value cost and government subsidy;

[0187] The algorithm optimization module is used for parameter adjustment and data storage based on the whale optimization algorithm, and applies the improved algorithm to path planning to obtain the path planning route and the global path length. The specific implementation of each module can refer to the description of the above method embodiment, and the embodiment of the present invention will not be repeated.

[0188] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for emergency material site selection and path optimization based on the whale optimization algorithm, characterized in that: The following steps are involved: Step 1: Establish an emergency material location selection model based on urgency and customer satisfaction, and select the optimal distribution point as the distribution center; Step 2: Establish an emergency material distribution model based on transportation cost and time, and plan the distribution route to the disaster site; Step 3: Adjust the parameters based on the whale algorithm and store the path generated by each iteration; Step 4: Obtain the path planning route of the disaster-affected point and the global path length.

2. According to claim 1, a method for emergency material site selection and path optimization based on a whale optimization algorithm is characterized in that: In step 1, the specific implementation process of establishing the emergency material site selection model is as follows: Step 1-1: Establish an urgency model; use the TOPSIS method combined with the entropy weight method to evaluate the urgency of emergency material needs at the disaster site; the specific implementation process is as follows: Step 1-1-1: Construct an initial decision matrix and perform standardization on it. The matrix composed of the index values ​​of n material demand points is called the initial decision matrix; Among them, Xij represents the index value of the index in the jth column of the i-th row of the material demand point, X is the initial decision matrix, i = 1, 2, .., n, j = 1, 2, .., m; Normalize X to avoid errors caused by different dimensions among indicators; Among them, C ij is the decision matrix, max(X j ) is the largest index value among the indexes in the jth column; Step 1-1-2: Determine the indicator weights and establish the weighted decision matrix R; Among them, R is the weighted decision matrix, which is obtained by multiplying the decision matrix and the weight obtained by the entropy weight method; Step 1-1-3: Determine the positive and negative ideal solutions; Positive ideal solution R + : R + =(R1 + ,R2 + ,..,R m + ),R j + =max(R ij ) Negative ideal solution R - : R - =(R1 - ,R2 - ,..,R m - ),R j - =min(R ij ) Among them, R + , R + is the set of positive and negative ideal solutions, represents the positive ideal solution of the mth column of the matrix, represents the negative ideal solution of the mth column of the matrix, represents the maximum positive ideal solution in the jth column of the weighted decision matrix, represents the minimum negative ideal solution in the jth column of the weighted decision matrix; Step 1-1-4: Calculate the positive and negative ideal solution distances; Among them, C + , C - Represent the distance of each solution to the positive and negative ideal solutions respectively; Step 1-1-5: Calculate the relative proximity of demand points; Among them, F1 represents the relative closeness of each disaster-affected point to the positive ideal solution, that is, the urgency; the larger the F1, the higher the urgency of the demand; Step 1-2: Establish a customer satisfaction model; including hard time windows and soft time windows. When emergency supplies do not arrive within the hard time window but arrive at the disaster site within the soft time window, customer satisfaction is calculated based on the degree of deviation from the hard time window. The customer satisfaction calculation formula is as follows: Among them, F2 is customer satisfaction, and the hard time window is (ET g ,LT g ), when the time range arrives, the customer satisfaction is 1; in the soft time window (et g ,lt g ) arrives outside the customer service center, customer satisfaction is 0; T g represents the time when the service vehicle arrives at the disaster site g; Step 1-3: Establish an emergency material model based on the urgency and customer satisfaction scores of each disaster site; normalize the urgency and satisfaction functions, and perform weight distribution to obtain the disaster site with the highest score, and select this site as the distribution center for rescue. The specific formula is as follows: Among them, F1 and F2 are urgency and satisfaction values, θ1 and θ2 are corresponding weights, and F is the comprehensive score.

3. The method for emergency material site selection and path optimization based on the whale optimization algorithm according to claim 2 is characterized in that: In step 2, the specific implementation process of establishing the emergency material distribution model is as follows: Step 2-1: Establish a transportation cost model based on risk management; transportation costs are reflected in the direct transportation costs related to transportation quantity, time, and distance during the transportation process, as well as the indirect transportation costs related to risk management; As the transportation time and distance increase, the risks faced by the goods during transportation will be higher; the heavier the goods, the greater the potential loss from accidents that may occur during transportation, and the risk management cost will increase accordingly; therefore, a transportation cost model based on risk management is established: Among them, C1 represents the transportation cost, d iq represents the distance from the i-th point to the q-th point; U represents the unit distance cost from the i-th point to the q-th point; G represents the quantity, quality or volume of the transported goods; R r is the risk area, R low is the low risk cutoff point, R mid is the medium risk cutoff point; C low is the fixed cost in the low risk interval; m mid and b mid are the slope and intercept of the medium risk interval; m high and b high are the slope and intercept of the high-risk interval; Step 2-2: Establish a vehicle travel time model; the time cost is used to describe the penalty caused by the vehicle allocation time exceeding the expected time due to road congestion; the shortest time for the vehicle to travel under unobstructed roads is taken as the expected time; based on this, the vehicle travel time model function is established: Among them, C2 represents the time cost, d iq is the distance from point i to point q; v iq is the speed; α is the cost coefficient of time delay; in order to quantitatively describe the driving conditions of vehicles under congested road conditions, R is defined as the "road traffic flow evaluation coefficient", and the road traffic conditions are divided into five levels: 0.00 < R < 0.30 indicates a smooth state; 0.30 < R < 0.60 indicates a slightly congested state; 0.60 < R < 0.75 indicates a moderately congested state; 0.75 < R < 0.90 indicates a relatively congested state; 0.90 < R < 1.00 indicates a congested state; R ≥ 1.00 indicates a severely congested state; Step 2-3: Establish a carbon cost model; the carbon emission cost is closely related to multiple variables in the transportation process, including the vehicle load, transportation distance and fuel consumption; Among them, C3 represents the carbon value cost, P * is the amount of fuel consumed per unit mileage; P0 is the amount of fuel consumed per unit mileage when the vehicle is not loaded with cargo; Q * is the maximum load of the vehicle; Q iq is the current vehicle load; Step 2-4: Establish a government subsidy model; during the distribution of emergency supplies, the government will provide corresponding subsidies based on the urgency of the situation; introducing the midpoint offset parameter μ into the formula can adjust the specific value of the urgency of supplies when the subsidy amount begins to increase significantly; Among them, S(P) represents the cost of government subsidies, expressed as C4, P is the urgency of materials, 0≤P≤1; k is the subsidy amplification factor, which is used to adjust the overall level of the subsidy amount; β is the slope parameter, which controls the rate at which the subsidy amount changes with P; the larger the β, the faster the change; the smaller the β, the smoother the change; μ is the midpoint offset parameter, which is used to adjust the position of the center point of the logistic function; when μ=0.5, the function is symmetric about P=0.5; when μ deviates from 0.5, the function will shift to the left or right; b is the base value or minimum value of the subsidy; Step 2-5: A total cost model is obtained based on transportation costs, vehicle travel time costs, carbon value costs and government subsidies. The calculation formula is as follows: C(cost)=C1+C2+C3-C4 Among them, C (cost) is the total cost; C1 is the transportation cost; C2 is the vehicle driving time cost; C3 is the carbon value cost; and C4 is the government subsidy cost.

4. According to claim 3, a method for emergency material site selection and path optimization based on a whale optimization algorithm is characterized in that: In step 3, the specific implementation process of introducing an improved whale optimization algorithm is as follows: Step 3-1: Improve the parameter a of the original Whale WOA algorithm, add a nonlinear convergence factor, and slow down the algorithm from falling into the local optimum as the number of iterations increases; at the same time, as the number of iterations gradually increases, the convergence speed is effectively accelerated; the changed algorithm parameter a is as follows: a=(a start -a end )+(1-t / t max ) / (1-μ1*t / t max ) Among them, a start and a end are the initial value and the final value of a respectively; μ1 is the adjustment coefficient, μ1=25, t max is the maximum number of iterations, t is the current number of iterations; Step 3-2: Introduce an adaptive search control coefficient to dynamically adjust the size of C according to the number of iterations; in the early stage of the algorithm, the C value is small, and the movement range of the whale increases, which helps to improve the global search ability; as the algorithm iterates, the movement range of the whale in the search space decreases, which helps to improve the local search ability; the expression is as follows: Among them, C is the adaptive search control coefficient, r is the coefficient that can control the decay rate of C, t max is the maximum number of iterations, t is the current number of iterations; Step 3-3: Introduce the memory list strategy; add a memory list to record the optimal solution encountered in each iteration; after reaching the maximum number of iterations, compare the optimal solution obtained in the last iteration with the solution in the memory, and select the better solution as the final result; avoid excessive memory capacity, limit the length of the memory list, and remove the worst solution when the maximum capacity is reached; the expression is as follows: S = remove(min(S)) Among them, each element stored in the S matrix is ​​the current optimal solution encountered during the iteration process. When the memory library reaches the maximum value, the minimum value in the library is removed by remove.

5. The method for emergency material site selection and path optimization based on the whale optimization algorithm according to claim 4 is characterized in that: In step 4, according to the improved whale optimization algorithm, the steps of obtaining the path planning route of the disaster-affected point and the global path length are specifically described as follows: Step 4-1: Set the parameters of the whale optimization ACSWOA algorithm: population size N, maximum number of iterations t max , the current iteration number t and the spiral shape constant b, and randomly initialize the initial position Xi of the whale group, where i = 1, 2, ..., n; Step 4-2: Calculate the fitness of each whale individual according to the following formula, sort the fitness, and find the whale individual corresponding to the minimum fitness, that is, the best whale individual X * And save it to the memory library list. When the library reaches the maximum capacity, remove the worst solution; S = remove(min(S)) Step 4-3: Calculate the parameter a that introduces the nonlinear convergence factor and update the coefficient vector A; a=(a start -a end )+(1-t / t max ) / (1-μ*t / t max ) Step 4-4: Introduce adaptive search control coefficients and update coefficient vector C; Step 4-5: Using α j =min(X i,j ),β j =max(X i,j ) Calculate the dynamic search boundary at this time; Step 4-6: Calculate the value of each whale individual, compare the solution of the reverse population with the solution in the current group, select s whale individuals with high fitness as the next generation group, and locate the whale individual with the minimum fitness as the prey position; Step 4-7: If t≤t max , then update the parameters a, A, C, l and p; Step 4-8: When p < 0.5; If |A|<1, update the spatial position of the current whale individual according to the following formula; If |A|>=1, then randomly select the whale individual position X rand , and update the spatial position of the current whale individual according to the following formula; D=|C·X rand (t)-X|,X(t+1)=X rand -A·D Among them, X*(t) is the position of the best solution currently obtained, X(t) is the current whale position, X(t+1) is the updated whale position, and D is the distance between the individual whale and the current optimal solution; Step 4-9: When p>=0.5, update the spatial position of the current individual whale according to the following formula; Where D' is the distance between the current search individual and the current optimal solution, l is a random number with a uniform distribution in the range [-1,1], and b is the spiral shape parameter; Step 4-10: Determine the new position according to the constraints. If the new position is feasible, the position of the individual whale will be updated. If not, the position of the individual whale will not change. Step 4-11: Determine the termination condition of the algorithm: If t = t max , then go to step 4-10; otherwise, repeat steps 4-7 to 4-11; Step 4-12: Output the spatial position X of the optimal whale individual * and its fitness, that is, output the emergency material distribution path and calculate the path length.

6. The method for emergency material site selection and path optimization based on the whale optimization algorithm according to claim 1 is characterized in that: The method is based on an emergency material site selection and path optimization system, which includes: The site selection and sorting module is used to sort the needs and satisfaction of different disaster-affected sites, and the disaster-affected site with the highest score is selected as the distribution center; The material distribution module is used to plan the material distribution route and obtain the low-cost and fast distribution route based on the transportation cost, vehicle travel time cost, carbon value cost and government subsidy; The algorithm optimization module is used for parameter adjustment and data storage based on the whale optimization algorithm, and applies the improved algorithm to path planning to obtain the path planning route and the global path length.

Images