Intensive storage network electric power material intelligent optimization scheduling method and system

Abstract

The invention discloses an intensive storage network electric power material intelligent optimization scheduling method and system, and belongs to the technical field of artificial intelligence, and the method comprises the steps: S1, building a database; s2, receiving logistics information, wherein the logistics information comprises article information or customized combination information and corresponding address information; s3, carrying out preliminary screening optimization on the transportation path by using an ant colony optimization algorithm based on an improved brainstorm algorithm; s4, classifying various vehicles corresponding to delivery vehicles, and formulating suitable vehicle types according to delivery requirements to receive orders; s5, path comprehensive optimization: providing path selection of reference cost based on a simulated annealing algorithm according to the vehicle oil consumption and the vehicle path, and finally combining the preliminary screening path in the S3 to obtain optimal path recommendation; according to the intensive storage network electric power material intelligent optimization scheduling method and system provided by the invention, a comprehensive reference sample is provided for establishing modernized and diversified logistics distribution path optimization, and the logistics distribution conditions of different types of electric power materials can be met.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a method and system for intelligent optimization and dispatching of electric power materials in an intensive storage network. Background Art

[0002] Intensive warehousing logistics refers to the realization of effective physical flow of various goods from the supply site to the receiving site within and outside the region, with large and medium-sized warehouses as the center, based on the scale and scope of regional warehousing, combined with the effective scope of logistics radiation. Logistics distribution is a key link in regional logistics operations and an implementation link for regional logistics to ultimately achieve logistics goals. In practice, the construction of logistics distribution path systems has problems such as a single optimization method and inconsistent data evaluation rules. How to use existing intelligent algorithms, big data analysis and data mining to achieve intelligent optimization design of the intensive warehousing network power material intelligent optimization dispatching system, and then intelligent management of the intensive warehousing network power material intelligent optimization dispatching system is an issue that needs to be solved urgently.

[0003] In addition, logistics distribution involves many issues such as vehicle wear and tear, fuel consumption, driver wages, transportation distance between multiple delivery points, and the number of delivery points. In these aspects, transportation distance and fuel consumption can be changed artificially, while the driver's salary and the number of distribution points are uncontrollable factors. The fuel consumption of the vehicle is mainly determined by the road and vehicle performance. Due to the change of vehicle performance, it cannot be used as a standard. Therefore, taking vehicle fuel consumption and vehicle path as the control objects at the same time, the reasonable dispatch of transportation vehicles is realized, and the high cost of logistics transportation is strived to be reduced. It provides a comprehensive reference sample for the establishment of a modern and diversified logistics distribution path optimization system, which can well meet the conditions of different types of power material logistics distribution. Summary of the invention

[0004] The purpose of the present invention is to provide a method and system for intelligent optimization scheduling of electric power materials in an intensive storage network to solve the problems existing in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides a method for intelligent optimization and dispatching of electric power materials in an intensive storage network, comprising the following steps:

[0006] S1. Establish a database, including the name and location information of each material warehouse, the fixed transportation routes between warehouses and the distribution vehicle parameters;

[0007] S2. Receive logistics information, including item information or customized combination information and corresponding address information;

[0008] S3, using the ant colony optimization algorithm based on the improved brainstorming algorithm to perform preliminary screening and optimization of the transportation path;

[0009] S4. Classify the various vehicles for delivery and formulate orders for the appropriate vehicle types according to the delivery requirements;

[0010] S5, comprehensive route optimization, provides route selection with reference cost based on simulated annealing algorithm according to vehicle fuel consumption and vehicle route, and finally obtains the optimal route recommendation by combining the preliminary screening path in S3.

[0011] Preferably, S3 obtains the route and the required number of vehicles K for serving all customers with the lowest total transportation cost based on the ant colony optimization algorithm (IBSO-ACO) of the improved brainstorming algorithm. The objective function of the vehicle routing problem is:

[0012]

[0013] Among them, c ij For v i to v j The total cost of the route, x ijk =1 means vehicle k starts from v i to v j Pass, otherwise x ijk =0,t ik Indicates that vehicle k arrives at customer warehouse v i time, and They are customer warehouse v i The earliest and latest time that the vehicle is required to arrive, when When the waiting cost per unit time is c w ;like The penalty cost per unit time is c p , c 0k is the dispatching cost of the distribution center; K is the number of vehicles required; c s is the travel cost per unit distance; s ij For warehouse v i To warehouse v j When i = 0, the more vehicles are needed, the greater the total cost, while minimizing the number of vehicles and the driving distance of all vehicles.

[0014] Preferably, some constraints of S3 based on the ant colony optimization algorithm of the improved brainstorming algorithm are:

[0015]

[0016] Constraint (1) means that each vehicle must start from distribution center v0 and return to distribution center v0 after delivery is completed; constraints (2) and (3) mean that each customer is served by only one vehicle; constraint (4) means that the load of the vehicle is less than its maximum carrying capacity; constraint (5) means that the arrival time is non-negative.

[0017] Preferably, the specific operation of the ant colony optimization algorithm based on the improved brainstorming algorithm in S3 is as follows:

[0018] Step 1: After completing a round of search for the vehicle routing problem based on the ant colony algorithm, M solutions are obtained; the solution of ant m is denoted as (S m , C m ), where S m is the total path and C m is the total cost. The improved brainstorming algorithm is used to update the solution set obtained by the ant colony algorithm. All the initial solution sets of vehicles obtained according to the basic ant colony algorithm are {(S m , C m )} m∈[1,M] ;

[0019] Step 2: Divide the solutions obtained in Step 1 into cluster A and cluster B and process them; cluster A contains all solutions with a cost less than (C min + C max ) / 2, and cluster B contains the remaining solutions. The optimal solution of cluster A is recorded as its cluster center, and a represents cluster A. The optimal solution of cluster B is similarly recorded as its cluster center, and b represents cluster B. For cluster A, each solution is replaced with a randomly generated solution; for cluster B, a randomly generated solution (S temp , C temp ) is generated, and then a random number r ∈ (0, 1) is generated. At the same time, a probability p is set in the algorithm. If r < p, the cluster center is set to If then it is replaced with (S temp , C temp ); if r ≥ p, a non-center solution is randomly selected as If then it is replaced with (S temp , C temp );

[0020] Step 3: Use the basic ant colony algorithm to regenerate new solutions;

[0021] Step 4: When the solutions obtained by the ant colony algorithm are updated by the improved brainstorming algorithm, the pheromone is updated in the next iteration to give positive feedback, and the optimal path is selected through multiple iterations.

[0022] Preferably, Step 4 is specifically as follows:

[0023] When the solution set obtained by the ant colony algorithm is updated by the improved brainstorming algorithm, the pheromone is updated in the next iteration to give positive feedback; the pheromone within the range τ is γ ij (τ), and the update rule is:

[0024]

[0025] Where ρ is the pheromone evaporation parameter; when the corresponding solution m is vehicle k, that is, m = k, From v i to v j The number of pheromones produced by vehicle k in the vicinity of τ on the route; if vehicle k passes through path e in range τ ij ,but otherwise The constant P represents the total amount of information. Its value affects the positive feedback of the algorithm, thereby affecting the convergence speed of the algorithm. k represents the total length of the path traversed by ant m in this iteration, Δγ ij (0)Initialized to 0.

[0026] The solution set updated at the iteration range τ contains both the optimal solution of the ant colony algorithm at iteration τ and the random solutions generated by the improved brainstorming algorithm, which enriches the diversity of pheromone distribution. The more pheromones a path has, the greater the probability that each vehicle will choose this path in the next iteration. Since pheromones also exist on a few randomly generated paths, they may also be selected by vehicles in the next iteration, which helps prevent local optimality. The ant colony optimization algorithm based on the improved brainstorming algorithm is iterated multiple times until the optimal cost converges or the maximum number of iterations is reached, and the optimal path is selected.

[0027] Preferably, various vehicles are classified in S4, including heavy trucks, medium trucks, light trucks and mini trucks.

[0028] Preferably, the path comprehensive optimization based on the simulated annealing algorithm in S5 specifically includes:

[0029] S51, initialization function, by generating an initial solution, using the function to objectively analyze the impact of the initialization node on the vehicle fuel consumption cost, and adjusting the number of iterations of the vehicle in the optimization process according to the initial solution;

[0030] S52, simulated annealing algorithm model, generates initial function value through new solution randomly generated by initialization function, sets Markov chain length internally based on temperature iteration factor, uses double-layer loop to globally traverse generated solution, selects three-point permutation method or two-point permutation method with random function, improves efficiency of annealing algorithm model;

[0031] S53, extracting the road gradient value and calculating the path cost; when calculating the cost distance matrix, adding the load coefficient and the road slope coefficient as the calculation cost;

[0032] S54, Metropolis criterion judgment, based on the comparison of the function value of the current solution with the new solution, when the function value of the new solution is lower, the new solution is accepted; when the function value of the new solution is worse (larger), the new solution is accepted with a certain probability.

[0033] Preferably, after completing the route assignment for a vehicle in S52, a total cost matrix is ​​obtained, a random function is set to execute using a two-point permutation method or a three-point permutation method of cost matrices i and j, and the obtained random solution is optimized based on the cost solution, and finally a comprehensive route recommendation is obtained.

[0034] Preferably, the criterion probability formula for Metropolis criterion judgment in S54 is:

[0035]

[0036] Among them, df = f(new) - f(old) represents the difference between the new solution and the old solution. When the new solution is greater than the old solution, the result is poor and is not selected. When the new solution is less than the old solution, it means that the value of the new solution is small, so the solution is directly accepted and the next step is traversed after the solution is obtained.

[0037] The initial route map is obtained through the initial logistics nodes, and then the vehicle load is weighted for each route, and the road slope is weighted; when passing a certain node, the vehicle's load capacity is reduced, and the fuel consumption of the next section is reduced. At the same time, the gradient data is extracted, and the adjustment of the gradient will also affect the analysis of fuel consumption.

[0038] The present invention also provides an intelligent optimization dispatching system for electric power materials in an intensive storage network, comprising a memory, a processor and a software program stored in the memory, wherein the software program executes all steps of the above method.

[0039] Therefore, the present invention adopts the above-mentioned intelligent optimization and dispatching method and system for electric power materials in an intensive warehousing network, which is used for optimizing the data and information status in the electric power material logistics distribution path optimization system, and realizes multi-dimensional information differentiation according to the differences and uniqueness of different path optimization schemes, and obtains and analyzes information of different dimensions. The intelligent management method of the traditional logistics distribution path optimization system is difficult to be scientifically guaranteed. Compared with the traditional logistics distribution path algorithm, the present invention performs multi-level data analysis on the scheme configuration strategy of the electric power material distribution path optimization system from multiple dimensions.

[0040] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a flow chart of an intelligent optimization dispatching method for intensive storage network power materials of the present invention;

[0042] Figure 2A flowchart of actual fuel consumption calculation provided by an embodiment of the present invention;

[0043] Figure 3 A flowchart of comprehensive path optimization provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0044] The following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0045] See also Figure 1 , a method for intelligent optimization and dispatching of electric power materials in an intensive storage network, comprising the following steps:

[0046] S1. Establish a database, including the name and location information of each material warehouse, the fixed transportation routes between warehouses and the distribution vehicle parameters;

[0047] S2. Receive logistics information, including item information or customized combination information and corresponding address information;

[0048] S3, using the ant colony optimization algorithm based on the improved brainstorming algorithm to perform preliminary screening and optimization of the transportation path;

[0049] A transportation system consisting of N customers, G = (V, E) is the transportation area map, where V is the set of all nodes including the distribution center node and the customer warehouse node, v0 represents the distribution center, v i , i∈[1,N] represents the target bin; E is the vehicle driving section. ij For v i to v j The total cost of the route, d i For the client warehouse node v i The demand, Q k is the maximum load that vehicle k can bear. ijk =1 means vehicle k starts from v i to v j Pass, otherwise x ijk =0. Customer v i The earliest and latest times that vehicles are required to arrive are and t ik Indicates that vehicle k arrives at customer warehouse v i When When the waiting cost per unit time is c w ;like The penalty cost per unit time is c p.

[0050] The ant colony optimization algorithm based on the improved brainstorming algorithm (IBSO-ACO) obtains the route with the lowest total transportation cost to serve all customers and the required number of vehicles K. The objective function of the vehicle routing problem is:

[0051]

[0052] Among them, c 0k is the dispatching cost of the distribution center; K is the number of vehicles required; c s is the travel cost per unit distance; s ij For warehouse v i To warehouse v j When i = 0, the more vehicles are needed, the greater the total cost. Considering the minimization of the number of vehicles and the driving distance of all vehicles, some constraints are:

[0053]

[0054] Constraint (1) means that each vehicle must start from distribution center v0 and return to distribution center v0 after delivery is completed; constraints (2) and (3) mean that each customer is served by only one vehicle; constraint (4) means that the load of the vehicle is less than its maximum carrying capacity; constraint (5) means that the arrival time is non-negative.

[0055] The specific operations of the ant colony optimization algorithm based on the improved brainstorming algorithm are:

[0056] Step 1: After completing a round of search for the vehicle routing problem based on the ant colony algorithm, M solutions are obtained; the solution of ant m is recorded as (S m ,C m ), where S m is the total path, C m is the total cost. The solution set obtained by the ant colony algorithm is updated using the improved brainstorming algorithm. The initial solution set of all vehicles obtained by the basic ant colony algorithm is {(S m ,C m )} m∈[1,M] ;

[0057] Step 2: Divide the solution obtained in step 1 into cluster A and cluster B and process them; cluster A contains the cost less than (C min +C max ) / 2, and cluster B contains the remaining solutions. The optimal solution of cluster A is The record is its cluster center, and a represents cluster A. The optimal solution for cluster B is The same record is recorded as its cluster center, and b represents cluster B. For cluster A, each solution is replaced with a randomly generated solution; for cluster B, a solution (S temp,C temp ), then generate a random number r ∈ (0, 1). At the same time, set the probability p in the algorithm. If r < p, the clustering center is set to If then replace it with (S temp ,C temp ); if r ≥ p, randomly select a non - central solution as If then replace it with (S temp ,C temp );

[0058] Step 3: Use the basic ant colony algorithm to regenerate a new solution;

[0059] Step 4: When the solution set obtained by the ant colony algorithm is updated by the improved brainstorming algorithm, update the pheromone in the next iteration to give positive feedback; the pheromone within the range τ is γ ij (τ), and the update rule is:

[0060]

[0061] Among them, ρ is the pheromone evaporation parameter; when the corresponding solution m is vehicle k, that is, m = k, is the amount of pheromone generated by vehicle k near τ on the route from v i to v j ; if vehicle k passes through path e ij in the range τ, then Otherwise The constant P represents the total amount of information, and its value affects the positive feedback of the algorithm, thus affecting the convergence speed of the algorithm. L k represents the total length of the path traveled by ant m in this iteration, and Δγ ij (0) is initialized to 0.

[0062] The solution set updated at the iteration range τ contains both the optimal solution of the ant colony algorithm at iteration τ and the random solutions generated by the improved brainstorming algorithm, enriching the diversity of pheromone distribution. The more pheromone on a path, the greater the probability that each vehicle will choose this path in the next iteration. Since pheromone also exists on a few randomly generated paths, they may also be selected by vehicles in the next iteration, which helps to prevent local optima. Through multiple iterations of the ant colony optimization algorithm based on the improved brainstorming algorithm until the optimal cost converges or reaches the maximum number of iterations, the optimal path is selected.

[0063] S4. For the corresponding delivery vehicles, classify various vehicles and formulate order - receiving suitable for vehicle types according to delivery requirements; including:

[0064] Step 1: Collect various logistics vehicle models in the logistics market;

[0065] Step 2: Divide into four categories according to type: heavy trucks, medium trucks, light trucks and mini trucks; the specific descriptions are as follows:

[0066] Heavy trucks are generally used to transport large equipment and non-detachable items, etc., and their general cargo capacity is usually more than 120 tons.

[0067] Medium-duty trucks usually carry about 50 tons of cargo and have almost the same efficiency as heavy-duty trucks.

[0068] Light trucks have a load capacity of about 26-35 tons and will not cause cargo loss after long-distance transportation.

[0069] Mini trucks generally have a load capacity of around 8-12 tons.

[0070] Step 3: Use light trucks and medium trucks as standard transportation tools. Assume that a distribution center has a total of k vehicles, r medium trucks, and l light trucks. The fuel consumption of a medium truck when fully loaded is Fuel consumption when no load is The fuel consumption of a light truck when fully loaded is Fuel consumption when no load is The distribution warehouse is V i (i∈(1,2,...,m)). When performing a delivery task, the load capacity of vehicle k is Q k , each vehicle in the distribution center departs from the distribution center and eventually returns to the distribution center. The distance between vehicle k and distribution center j is The average gradient between two delivery points is The gradient from the distribution center to the distribution point is The fuel consumption coefficient under different gradients is σ. The fuel consumption of the vehicle is linearly related to the load. Figure 2 This is the actual fuel consumption calculation process based on the load balancing fuel consumption method.

[0071] Step 4: Use the vehicle's load condition to estimate the fuel consumption rate, and there is a linear relationship between the actual fuel consumption rate and the vehicle's load. In order to be closer to reality, a constant multiplication is performed after the load calculation. The following formula is the actual fuel consumption rate calculation formula for trucks:

[0072]

[0073] Among them, F(q) represents the actual fuel consumption rate (L / 100km), q is the actual load during driving, Q is full load, c is the fuel consumption correction coefficient (randomly ranging from 0.5 to 1), corresponding to the actual weight of the cargo, is the fuel consumption of the vehicle when it departs from the distribution center, The fuel consumption of the vehicle throughout the journey.

[0074] When the vehicle is fully loaded, the fuel consumption is When the vehicle is unloaded, the fuel consumption is Gek*c. When the vehicle is loaded but not fully loaded, the fuel consumption cost from cargo compartment i to cargo compartment j is

[0075] In terms of road slope, the greater the road slope, the higher the vehicle's fuel consumption. The fuel consumption corresponding to each longitudinal slope during driving should be somewhat different from that during normal driving. The total fuel consumption value calculated based on the slope should be:

[0076]

[0077] Among them, i is the gradient, F(q) is the actual fuel consumption rate, Y diff is the difference between the longitudinal slope fuel consumption and the normal driving fuel consumption. If the transport vehicle is from i to j, then Fuel consumption On the contrary Some parameter constraints are as follows:

[0078]

[0079] S5, comprehensive route optimization, based on vehicle fuel consumption and vehicle route, provides route selection with reference cost based on simulated annealing algorithm, and finally obtains the optimal route recommendation by combining the preliminary screening route in S3, including:

[0080] S51, initialization function, by generating an initial solution, using the function to objectively analyze the impact of the initialization node on the vehicle fuel consumption cost, and adjusting the number of iterations of the vehicle in the optimization process according to the initial solution;

[0081] S52, simulated annealing algorithm model, generates initial function value through new solution randomly generated by initialization function, sets Markov chain length internally based on temperature iteration factor, uses double-layer loop to globally traverse generated solution, selects three-point permutation method or two-point permutation method with random function, improves efficiency of annealing algorithm model;

[0082] S53, extracting the road gradient value and calculating the path cost; when calculating the cost distance matrix, adding the load coefficient and the road slope coefficient as the calculation cost;

[0083] S54, Metropolis criterion judgment, based on the comparison of the function value of the current solution with the new solution, when the function value of the new solution is lower, the new solution is accepted; when the function value of the new solution is worse (larger), the new solution is accepted with a certain probability.

[0084] The probability formula is:

[0085]

[0086] Among them, df = f(new) - f(old) represents the difference between the new solution and the old solution. When the new solution is greater than the old solution, the result is poor and is not selected. When the new solution is less than the old solution, it means that the value of the new solution is small, so the solution is directly accepted and the next step is traversed after the solution is obtained.

[0087] The initial route map is obtained through the initial logistics nodes, and then the vehicle load is weighted for each route, and the road slope is weighted; when passing a certain node, the vehicle's load capacity is reduced and the fuel consumption of the next section is reduced. At the same time, the gradient data is extracted, and the adjustment of the gradient will also affect the analysis of fuel consumption, such as Figure 3 shown.

[0088] In this embodiment, the data nodes collected on the data platform are used for analysis. Based on the initial position information, the initial distance matrix is ​​obtained. The average gradient between two distribution points is calculated using the gradient matrix. In the actual logistics transportation process, the transportation tasks of each vehicle are fixed, so the transportation demand of the truck at each demand point is fixed; at the same time, the load of the vehicle during transportation will also change. The cost of a single vehicle under the static distribution between the total load capacity and the distribution points is:

[0089]

[0090] in, is the distance between vehicle k and two points i and j, and i is the gradient value. Considering the load of the vehicle, the transportation cost of a single vehicle between two points i and j can also be solved as:

[0091]

[0092] The real-time fuel consumption difference is calculated by calculating the ratio of the load difference between i and j to the full load and the fuel consumption difference between i and j of vehicle k, thereby accurately calculating the fuel consumption of the vehicle.

[0093] The optimal path recommendation obtained comprehensively adopts the three-point permutation method or the two-point permutation method. On the basis of the probability theory method, a specific exchange method is used to determine the neighborhood operation, improve the randomness of the new solution space, and reduce the algorithm from falling into the local optimal problem. After completing the assignment of the path to a certain vehicle, the total cost matrix is ​​obtained. The two-point permutation method or the three-point permutation method of the cost matrix (i and j) is used to set a random function execution, and the obtained random solution is optimized based on the cost solution, and finally a comprehensive path recommendation is obtained.

[0094] An intelligent optimization dispatching system for electric power materials in an intensive storage network includes a memory, a processor, and a software program stored in the memory, and the software program executes all steps of the above method.

[0095] Therefore, the present invention adopts the above-mentioned intelligent optimization and scheduling method and system for electric power materials in an intensive warehousing network, based on a variety of intelligent algorithms, for optimizing the data and information status in the electric power material logistics distribution path optimization system, and performs multi-level data analysis on the solution configuration strategy of the electric power material distribution path optimization system from multiple dimensions, providing a comprehensive reference sample for establishing a modern and diversified logistics distribution path optimization system, which can well meet the conditions of different types of electric power material logistics distribution.

[0096] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A method for intelligent optimization and dispatching of power materials in an intensive storage network, characterized in that: The following steps are involved: S1. Establish a database, including the name and location information of each material warehouse, the fixed transportation routes between warehouses and the distribution vehicle parameters; S2. Receive logistics information, including item information or customized combination information and corresponding address information; S3, using the ant colony optimization algorithm based on the improved brainstorming algorithm to perform preliminary screening and optimization of the transportation path; S4. Classify the various vehicles for delivery and formulate orders for the appropriate vehicle types according to the delivery requirements; S5, comprehensive path optimization, provides path selection with reference cost based on simulated annealing algorithm according to vehicle fuel consumption and vehicle path, and finally obtains the optimal path recommendation by combining the preliminary screening path in S3.

2. According to claim 1, a method for intelligent optimization and dispatching of power materials in an intensive storage network is characterized in that: S3 uses the ant colony optimization algorithm based on the improved brainstorming algorithm to obtain the route with the lowest total transportation cost to serve all customers and the required number of vehicles K. The objective function of the vehicle routing problem is: Among them, c ij For v i to v j The total cost of the route, x ijk =1 means vehicle k starts from v i to v j Pass, otherwise x ijk =0,t ik Indicates that vehicle k arrives at customer warehouse v i time, and They are customer warehouse v i The earliest and latest time that the vehicle is required to arrive, when When the waiting cost per unit time is c w ;like The penalty cost per unit time is c p , c 0k is the dispatching cost of the distribution center; K is the number of vehicles required; c s is the travel cost per unit distance; s ij For warehouse v i To warehouse v j distance.

3. According to claim 2, a method for intelligent optimization and dispatching of power materials in an intensive storage network is characterized in that: The constraints of the ant colony optimization algorithm based on the improved brainstorming algorithm S3 are: Constraint (1) means that each vehicle must depart from distribution center v0 and return to distribution center v0 after delivery is completed; constraints (2) and (3) mean that each customer is served by one vehicle; constraint (4) means that the load of the vehicle is less than its maximum carrying capacity; constraint (5) means that the arrival time is non-negative.

4. According to claim 1, a method for intelligent optimization and dispatching of power materials in an intensive storage network is characterized in that: The specific operation of S3's ant colony optimization algorithm based on the improved brainstorming algorithm is: Step 1: After completing a round of search for the vehicle routing problem based on the ant colony algorithm, M solutions are obtained; Step 2: Divide the solution obtained in step 1 into cluster A and cluster B and process them; Step 3: Use the basic ant colony algorithm to regenerate a new solution; Step 4: After the solution obtained by the ant colony algorithm is updated by the improved brainstorming algorithm, the pheromone is updated in the next iteration to give positive feedback, and the optimal path is selected through multiple iterations.

5. According to claim 4, a method for intelligent optimization and dispatching of power materials in an intensive storage network is characterized in that: Step 4 is as follows: When the solution set obtained by the ant colony algorithm is updated by the improved brainstorming algorithm, the next iteration updates the pheromone to give positive feedback; the pheromone within the range τ is γ ij (τ), the update rule is: Where ρ is the pheromone evaporation parameter; when the corresponding solution m is vehicle k, that is, m = k, From v i to v j The number of pheromones produced by vehicle k in the vicinity of τ on the route; if vehicle k passes through path e in range τ ij ,but otherwise The constant P represents the total amount of information, L k represents the total length of the path traversed by ant m in this iteration, Δγ ij (0)Initialized to 0. The solution set updated at the iteration range τ includes both the optimal solution of the ant colony algorithm at iteration τ and the random solution generated by the improved brainstorming algorithm. The ant colony optimization algorithm based on the improved brainstorming algorithm is iterated multiple times until the optimal cost converges or the maximum number of iterations is reached, and the optimal path is selected.

6. The method for intelligent optimization and dispatching of electric power materials in an intensive storage network according to claim 1, characterized in that: Various types of vehicles are classified in S4, including heavy trucks, medium trucks, light trucks and mini trucks.

7. The method for intelligent optimization and dispatching of electric power materials in an intensive storage network according to claim 1, characterized in that: The path comprehensive optimization based on simulated annealing algorithm in S5 specifically includes: S51, initialization function, by generating an initial solution, using the function to objectively analyze the impact of the initialization node on the vehicle fuel consumption cost, and adjusting the number of iterations of the vehicle in the optimization process according to the initial solution; S52, simulated annealing algorithm model, generates initial function value through new solution randomly generated by initialization function, sets Markov chain length internally based on temperature iteration factor, uses double-layer loop to globally traverse generated solution, and selects three-point permutation method or two-point permutation method with random function; S53, extracting the road gradient value and calculating the path cost; when calculating the cost distance matrix, adding the load coefficient and the road slope coefficient as the calculation cost; S54, Metropolis criterion judgment, based on the comparison of the function value of the current solution with the new solution, when the function value of the new solution is low, the new solution is accepted, and when the function value of the new solution is large, the new solution is accepted with a certain probability.

8. The method for intelligent optimization and dispatching of electric power materials in an intensive storage network according to claim 7, characterized in that: After the path is assigned to a vehicle in S52, the total cost matrix is ​​obtained, and a random function is set to execute using the two-point permutation method or the three-point permutation method of the cost matrices i and j, and the obtained random solution is optimized based on the cost solution, and finally a comprehensive path recommendation is obtained.

9. The method for intelligent optimization and dispatching of electric power materials in an intensive storage network according to claim 7, characterized in that: The criterion probability formula for Metropolis criterion judgment in S54 is: Among them, df = f(new) - f(old) represents the difference between the new solution and the old solution. When the new solution is greater than the old solution, the result is poor and is not selected. When the new solution is less than the old solution, it means that the value of the new solution is small, so the solution is directly accepted and the next step is traversed after the solution is obtained.

10. A system for intelligent optimization and dispatching method of power materials in intensive storage network according to any one of claims 1 to 9, characterized in that: The system comprises a memory, a processor and a software program stored in the memory, wherein the software program executes all the steps of the above method.

Images