Medical material automatic guided vehicle distribution path optimization system and method

Abstract

The invention relates to the technical field of guide vehicle distribution, in particular to a medical material automatic guide vehicle distribution path optimization system and method, which are used for investigating and analyzing hospital layout, and understanding and analyzing personnel, traffic flow, medical material characteristics and distribution requirements; selecting a proper AGV type and navigation mode according to an analysis result; minimization of distribution time is determined, and AGV energy consumption and equipment loss cost are reduced; the method comprises the following steps: establishing a mathematical model, abstracting a hospital environment into a directed graph, constructing a shortest path model or a multi-objective optimization model, and considering constraint conditions including a time window and a priority; an improved genetic algorithm is adopted to solve the model, the distribution path of the AGV is encoded into chromosome, fitness function design, selection, intersection and mutation operation are performed, the optimal distribution path is searched, the distribution path of the medical materials is optimized through model construction and algorithm improvement, the distribution time is shortened, and the distribution efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the field of guided vehicle distribution technology, and in particular to a system and method for optimizing the distribution path of an automatic guided vehicle for medical supplies. Background Art

[0002] In the past, medical supplies were usually delivered manually, but traditional manual delivery methods are easily affected by limitations such as staff fatigue and working hours.

[0003] As the importance of AGV in the distribution of medical supplies continues to increase, in order to improve the distribution efficiency, AGV (Automated Guided Vehicle) as an automated logistics equipment can navigate autonomously in the medical environment. Compared with the traditional manual distribution method, AGV is not restricted by personnel fatigue, working hours, etc., and can perform distribution tasks continuously and stably. It can quickly and accurately transport medical supplies from the storage area to various use points, such as wards, operating rooms, laboratories, etc., greatly shortening the distribution time and ensuring the timeliness of medical services. Improve the accuracy of distribution. AGV relies on advanced navigation and control systems to accurately travel along the preset path. In the process of distributing medical supplies, it can avoid distribution errors caused by human factors, such as sending to the wrong department, sending the wrong type or quantity of materials, etc., improve the accuracy and quality of distribution, and ensure the smooth development of medical work. It can also ensure the safety of the medical environment. In an environment such as a hospital with extremely high requirements for hygiene and safety, AGV can reduce personnel flow and contact, and reduce the risk of cross infection. At the same time, its stable operating performance also reduces the damage to materials or accidental injuries to surrounding personnel that may be caused by manual handling, providing a safer guarantee for the medical environment.

[0004] However, due to the limitations of AGV's own navigation accuracy and error, speed and safety, it is difficult for AGV to classify and distribute diverse materials along complex paths in complex medical environments. Summary of the invention

[0005] The purpose of the present invention is to provide a medical supplies automatic guided vehicle distribution path optimization system and method, aiming to solve the problem that AGV is difficult to accurately classify and distribute diverse materials in a complex path in a complex medical environment due to its own navigation accuracy and error and speed and safety limitations.

[0006] To achieve the above object, the present invention provides a method for optimizing the distribution path of an automated guided vehicle for medical supplies, comprising the following steps:

[0007] Conduct research and analysis on hospital layout, understand and analyze personnel and traffic flows, as well as the characteristics and distribution needs of medical supplies;

[0008] According to the analysis results, select the appropriate AGV type and navigation method;

[0009] Determine the minimum delivery time and reduce AGV energy consumption and equipment loss costs;

[0010] Establish a mathematical model, abstract the hospital environment into a directed graph, construct a shortest path model or a multi-objective optimization model, and consider constraints, including time windows and priorities;

[0011] An improved genetic algorithm is used to solve the model, and the optimal delivery path is found by encoding the AGV's delivery path into chromosomes, designing fitness functions, and performing selection, crossover and mutation operations.

[0012] Among them, the hospital layout is investigated and analyzed, and the personnel and traffic flow as well as the characteristics and distribution needs of medical supplies are understood and analyzed. The steps also include:

[0013] The hospital layout includes floor distribution, department location, passage width and connection, and a detailed layout diagram.

[0014] Among them, the hospital layout is investigated and analyzed, and the personnel and traffic flow as well as the characteristics and distribution needs of medical supplies are understood and analyzed. The steps also include:

[0015] Classify medical supplies, analyze the weight, volume, storage conditions, delivery time and priority requirements of each type of supplies, and convert these requirements into constraints for path optimization.

[0016] The improved genetic algorithm is used to solve the model, and the optimal delivery path is found through encoding, fitness function design, selection, crossover and mutation operations. The steps also include:

[0017] The genetic algorithm includes adopting a path-based encoding method, designing a fitness function, comprehensively considering delivery time, cost and load balance, and converting multi-objectives into single-objective problems in a weighted manner for solution.

[0018] The improved genetic algorithm is used to solve the model, and the optimal delivery path is found through encoding, fitness function design, selection, crossover and mutation operations. The steps also include:

[0019] Computer simulation software is used to build a virtual scene similar to the actual hospital environment, simulate different medical supplies distribution tasks and various interference factors, and verify the effectiveness and feasibility of the path optimization model and algorithm.

[0020] A medical material automatic guided vehicle distribution path optimization system applies the medical material automatic guided vehicle distribution path optimization method.

[0021] The present invention provides a medical supplies automatic guided vehicle distribution path optimization system and method, which optimizes the distribution path of medical supplies through model construction and algorithm improvement, thereby shortening the distribution time and improving the distribution efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art are briefly introduced below.

[0023] Figure 1 It is a step diagram of the method for optimizing the distribution path of medical supplies automatic guided vehicles according to the first embodiment of the present invention.

[0024] Figure 2 It is a genetic algorithm flow chart of the method for optimizing the distribution path of medical supplies automatic guided vehicles according to the first embodiment of the present invention.

[0025] Figure 3 It is a schematic diagram of the distribution center task allocation process of the medical supplies automatic guided vehicle distribution path optimization method according to the first embodiment of the present invention.

[0026] Figure 4 It is a schematic diagram of the vehicle task allocation process of the medical supplies automatic guided vehicle distribution path optimization method according to the first embodiment of the present invention.

[0027] Figure 5 It is a chromosome crossover flow chart of the method for optimizing the distribution path of medical supplies automatic guided vehicles according to the first embodiment of the present invention.

[0028] Figure 6 It is a chromosome variation flow chart of the method for optimizing the distribution path of medical supplies automatic guided vehicles according to the first embodiment of the present invention.

[0029] Figure 7 It is a schematic diagram of the improved genetic algorithm flow of the medical supplies automatic guided vehicle distribution path optimization method according to the first embodiment of the present invention. DETAILED DESCRIPTION

[0030] Embodiments of the present invention are described in detail below. Examples of the embodiments are shown in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, but should not be construed as limiting the present invention.

[0031] The first embodiment of the present application is:

[0032] See also Figure 1 ,in Figure 1 It is a step diagram of the method for optimizing the distribution path of medical supplies automatic guided vehicles according to the first embodiment of the present invention.

[0033] The present invention provides a method for optimizing the distribution path of an automated guided vehicle for medical supplies, comprising the following steps:

[0034] S101: Investigate and analyze hospital layout, understand and analyze personnel and traffic flows, as well as the characteristics and distribution needs of medical supplies;

[0035] S102: Selecting a suitable AGV type and navigation method according to the analysis results;

[0036] S103: Determine the minimization of delivery time and the reduction of AGV energy consumption and equipment loss costs;

[0037] S104: Establish a mathematical model, abstract the hospital environment into a directed graph, construct a shortest path model or a multi-objective optimization model, and consider constraints, including time windows and priorities;

[0038] S105: The model is solved using an improved genetic algorithm, and the optimal delivery path is found by encoding the delivery path of the AGV into chromosomes, fitness function design, selection, crossover and mutation operations.

[0039] Specifically, statistics are collected on the daily delivery quantity of various types of medical supplies. For example, emergency medicines need to be delivered 20-30 times on average every day, and the quantity each time depends on different emergency scenarios; the number of ordinary medicines delivered is about 200-300 times, involving a variety of dosage forms and specifications; the delivery volume of medical devices is about 50-80 pieces, including large equipment, small precision instruments, etc.; the daily collection and delivery volume of test samples is 500-800 pieces, and the delivery task information of materials from starting points such as pharmacies and warehouses to various departments, wards, operating rooms, and testing centers is clarified. For example, the operating room has a frequent demand for emergency medicines and high-value medical devices, and the starting point of its delivery may be the central pharmacy or the equipment warehouse, and the end point is the operating room on different floors; the starting point of ordinary medicine delivery in the ward is the ward pharmacy, and the end point is each ward. Information such as the number of floors, area of ​​each floor, and location of stairs and elevators in each building of the hospital is obtained. For example, the main building of the hospital has 10 floors, each floor has an area of ​​about 2,000 square meters, with 4 elevators and 2 stairwells. The length, width, slope and other parameters of the passage, as well as the direction of passage (one-way or two-way) were measured in detail. For example, the main passage connecting the ward building and the medical technology building is 100 meters long and 3 meters wide, and is a two-way passage; while the passages inside some departments are one-way and only 1.5 meters wide. At the same time, crowded areas (such as outpatient halls, waiting areas, etc.) and places prone to congestion (such as near elevator entrances) were marked, and the maximum driving speed of AGV (generally 1-2 meters / second), acceleration, turning radius (such as 0.8-1.2 meters) and other parameters were mastered. Different types of AGVs vary in speed and load capacity. For example, backpack AGVs are slower but have moderate load capacity and can be used to deliver heavier medical devices. Towing AGVs are faster and suitable for bulk delivery of common medicines. The AGV navigation system (such as improved laser navigation) was tested for accuracy. Under ideal conditions, the positioning accuracy can reach ±5 mm, but under interference such as strong light and complex electromagnetic environments, the accuracy may drop to ±10-15 mm. The frequency and location of navigation errors were recorded so that error compensation can be considered in path planning.

[0040] Based on the above data, a multi-objective optimization path planning model was constructed. The optimization objectives are the shortest delivery time, the lowest cost (including energy consumption cost and equipment loss cost) and load balance. For example, the delivery time objective function can be expressed as, where is the driving time of the AGV from node to node, is the decision variable, indicating whether to pass the edge; the cost objective function takes into account the energy consumption and equipment depreciation cost of the AGV, which is related to the driving distance and time; the load balance objective is measured by calculating the load variance of each AGV, and the improved genetic algorithm is used for solution. The encoding method of the traditional genetic algorithm is improved, and the path-based encoding method is adopted to encode the AGV's delivery path as a chromosome. A suitable fitness function is designed, which comprehensively considers the three optimization objectives, and converts the multi-objective into a single-objective problem by weighting. For example, the fitness function, where are the weights of the delivery time, cost, and load balance objectives respectively. By adjusting the weights, the importance of different objectives can be balanced. In the solution process, the genetic algorithm parameters such as population size, crossover probability, and mutation probability are set. After multiple iterations (for example, 500-1000 times), a set of better delivery path solutions are obtained. The results show that compared with the initial random path, the optimized delivery time is shortened by an average of about 35%, the cost is reduced by about 25%, and the load variance between AGVs is reduced by about 40%, indicating that the model and solution algorithm have significant effects in improving distribution efficiency and balancing load. Through the construction of the model and the improvement of the algorithm, the distribution path of medical supplies is optimized, thereby shortening the distribution time and improving the distribution efficiency.

[0041] Furthermore, parameter sensitivity analysis aims to study the degree of influence of changes in various parameters in the model on the optimization results (delivery time, cost, load balance). Through analysis, it can be determined which parameters have a greater impact on the results and which parameters are relatively less sensitive, thereby providing a basis for further optimization of the model and parameter adjustment in practical applications. Single parameter change analysis: Change the parameters in the model one by one, such as channel weight (weight considering factors such as distance and personnel flow), AGV speed, load capacity, material priority weight, etc., and only change one parameter at a time to observe the changes in the optimization results. For example, when the weight of a densely populated channel is increased by a certain proportion, the changes in delivery time and cost are analyzed. At the same time, multiple parameters are changed to study the impact of their interaction on the results. Methods such as Latin hypercube sampling are used to generate parameter combinations, and statistical methods such as regression analysis are used to analyze the interaction effects between parameters. Sensitive parameters: It is found that the material priority weight has a significant impact on the delivery time of emergency drugs and high-value medical devices. When the priority weight of emergency materials is increased, its delivery time is significantly shortened, but it may cause a slight increase in the delivery time of ordinary materials. The speed parameter of AGV has a great influence on delivery time and cost. Increasing the speed can shorten the time but may increase energy consumption costs. The weight of the personnel flow factor in the channel weight is also sensitive to path selection and delivery time. Reasonable adjustment of the weight of the crowded channel can effectively optimize the path. Insensitive parameters: some parameters have little effect on the results within a certain range. For example, the minimum turning radius of AGV has little effect on the overall delivery efficiency and cost under the existing channel width of the hospital; the delivery time window of some common drugs is relaxed or tightened within a certain range, which has no significant impact on the optimization results. Through data description, model solution and parameter sensitivity analysis, it provides comprehensive theoretical support and practical guidance for the optimization of the AGV distribution path of medical supplies in this large comprehensive hospital, which is helpful to further improve the distribution plan and improve the quality of medical services. The optimized path can greatly shorten the delivery time of emergency medical supplies, which is shortened by an average of [X]%, effectively meeting the requirements for timeliness of supplies in emergency scenarios.

[0042] For further information, see Figure 2 , genetic algorithm achieves the continuous evolution of chromosomes by imitating the natural selection and chromosome replication, crossover, gene mutation and other behaviors in the biological evolution process, and tries to find the global optimal solution to the complex problem with the maximum fitness through iteration. Its principle and coding are relatively simple, and it can be applied to both discrete and continuous problems. Therefore, it has always shown strong vitality and has been continuously expanded by later scholars and widely used in path planning, image processing, site selection and layout, production scheduling and other fields. The implementation of genetic algorithm includes five steps: setting encoding and decoding logic, population initialization, selection according to fitness function, crossover and mutation, and iteration to termination condition.

[0043] (1) Setting up encoding and decoding logic. Common encoding methods used in genetic algorithms include binary encoding, natural number encoding, and real number encoding. Setting up encoding and decoding logic is the basis for combining chromosomes with actual problems. The actual problem to be solved is converted into a string of data that is easy for the computer to process through encoding, and then the specific parameters of the actual problem solution represented by the chromosome are extracted from the data string through decoding.

[0044] (2) Population initialization. Before selecting and cultivating a population, a population must first be created through population initialization. Common initialization methods include random generation, fixed value setting, two-step method, specific application method, etc. The quality of the initial population has a great influence on the search range and convergence speed.

[0045] (3) Selection based on the fitness function. This step is the key to the evolution of the population. The fitness function represents the quality of the chromosome. The higher the function value, the closer the solution represented by the chromosome is to the goal that the algorithm wants to achieve. Selection is a concrete manifestation of the concept of "survival of the fittest" in Darwin's theory of evolution. According to the fitness of the individual, inferior individuals are eliminated with a certain probability, thereby achieving the evolution of the entire population. Common selection strategies include roulette selection and tournament selection.

[0046] (4) Crossover and mutation. Crossover and mutation generate new individuals by imitating the crossover of chromosomes and gene mutation in the process of biological reproduction, allowing the algorithm to continuously search for new feasible domains instead of simply screening and sorting the initial population. Crossover can be either self-pollination or hybridization. By exchanging fragments of parent chromosomes, the excellent fragments of the parent can be combined together. Mutation is a random perturbation of the gene at a certain gene position, allowing the algorithm to jump out of the local optimum.

[0047] (5) Iterate until the termination condition. Optimization is carried out in the continuous iteration of the algorithm, and a termination condition needs to be set to stop the iteration. The termination condition can be either the objective function reaching a certain value or the population iteration reaching a specified number of times. The termination condition should be appropriate. If the algorithm is terminated early, the iteration will be stopped when there is still a certain distance from the optimal solution. If it is terminated too late, it will take too long and cause a waste of resources.

[0048] For further information, see Figure 3 and Figure 4, Considering that roulette selection has a certain degree of randomness, the probability of the optimal solution being selected is relatively high, but it may still be lost in the inheritance process or destroyed in evolution. Therefore, an elite retention strategy is introduced in the genetic algorithm to put the best individuals of each generation directly into the offspring without selection and evolution operations to ensure the survival of the optimal solution. In addition, although the objectives considered in this paper are complex, involving many factors such as distance, carbon emissions, and time windows, the objective function is generally highly correlated with the distance. Short delivery distances not only directly lead to reduced energy consumption and transportation costs, but also reduce carbon emissions, cargo damage, and refrigeration costs. Therefore, nodes can be initially allocated according to distance to speed up the convergence of the algorithm.

[0049] Before applying the improved genetic algorithm with elite retention strategy to the multi-center cold chain distribution problem under carbon trading, we must first consider the encoding and decoding methods of chromosomes for subsequent crossover, mutation, selection and other operations. The vehicle routing problem studied in this paper is aimed at the permutation and combination order of customer nodes, so natural number encoding is used. For the problem with |N| customers and |M| distribution centers, the chromosome length is defined as |M|+|N|-1, and numbers 0 to |N|-1 are customer nodes. Numbers greater than |N|-1 are regarded as the segmentation points of different distribution centers. According to the encoding idea, the following operations need to be performed during decoding:

[0050] (1) Traverse the chromosome nodes and sequentially place the nodes with numbers less than or equal to |N|-1 into distribution center m. When encountering a node with a number greater than |N|-1, skip it and assign the subsequent node to distribution center m+1.

[0051] (2) The tasks assigned to each distribution center are assigned to specific vehicles according to the load limit.

[0052] For further information, see Figure 6 ,Considering that the objective function is highly correlated with the delivery distance, the distribution ,method in which the distribution center and the customer nodes it is responsible for are generally far from each ,distribution method lacks discussion value. ,In order to improve the convergence speed, the nodes are ,preliminarily allocated first, and each customer point is ,assigned to the nearest distribution center by distance, and then the tasks of each distribution ,center are randomly arranged, and a number greater than |N|-1 is ,inserted between the tasks to form a complete chromosome.

[0053] Fitness is the only basis for judging the quality of chromosomes during the operation of the algorithm. For individuals, the larger the fitness value, the higher the adaptability to the environment, and the easier it is to be selected to survive and reproduce, thus producing offspring. By decoding the chromosome, the total cost of the distribution plan it represents can be calculated. In the multi-center cold chain distribution problem under carbon trading, the lower the total cost, the better the individual. Therefore, the difference between the maximum cost in the current population and chromosome i is taken as its fitness value:

[0054] fitness i = maxC - C i , i ∈ {1, 2, ..., popsize}

[0055] In the above formula, maxC represents the minimum value of the total cost corresponding to all chromosomes in the population, C i represents the cost corresponding to chromosome i, and popsize represents the population size. By taking the difference from the maximum total cost, the smaller the objective function, the greater the retention probability, while preventing the difference between individual fitness values from being too small to effectively screen. During the execution process, first calculate the total cost corresponding to all chromosomes, select the maximum value among them and record it as maxC, and then traverse the population to calculate the fitness values of all individuals in the population.

[0056] The selection operator is the key to ensuring the continuous evolution of the population and making it more and more adaptable to the environment as the number of iterations increases. Analyze two commonly used selection operators. Roulette wheel makes the probability of each individual being selected proportional to its fitness value by simulating a roulette wheel. Chromosomes, regardless of their quality, may survive, only with different probabilities; tournament randomly selects a certain number of individuals to participate in the competition each time, and selects the best individuals among them. The individuals with the lowest fitness values in the entire parental population will definitely be eliminated. The problem studied in this paper is a discrete problem. Exchanging two nodes may cause a large change in cost, that is, a poor solution may still have the potential to be transformed into a good solution after genetic mutation. Therefore, it is hoped that a poor solution still has a certain probability of being accepted, and roulette wheel is used as the selection strategy. The implementation steps of roulette wheel in the algorithm are as follows:

[0057] (1) Decode the chromosome to obtain the objective function value, and calculate the fitness value fitness of each chromosome according to the fitness function i ;

[0058] (2) Normalize the fitness set, calculate the survival probability of each chromosome, and the probability that the i-th chromosome is retained is:

[0059] (3) Construct a cumulative probability list, and the i-th item in the list

[0060] (4) Generate a random number r between 0 and 1, select the chromosome according to its landing point. When 0 < r < q1, take out the first chromosome. When q i-1 < r < q i (i > 1), then take out the i-th chromosome and put it into the parent. Repeat this step popsize - 1 times to obtain a parent set without elites.

[0061] The crossover operator is one of the two ways to generate new individuals in genetic algorithms. The gene fragments of the parent chromosomes are crossed according to certain rules. While inheriting the excellent genes of the parent, a better combination method is searched to reproduce offspring with higher fitness. According to the number of parents required to produce offspring, there are three types: asexual, sexual, and multi-parent recombination. In the multi-center cold chain distribution problem, each gene appears and can only appear once. The fragments separated by the split point represent the tasks assigned to different distribution centers. Therefore, it is more suitable to use asexual crossover. The crossover between the chromosome fragments representing the tasks of each distribution center is adjusted to optimize the task allocation plan between distribution centers, avoiding the problem of repeated genes caused by sexual and multi-parent recombination. According to the gene extraction method, the crossover methods include single-point crossover, multi-point crossover, uniform crossover, etc. In view of the characteristics of multi-center distribution, this paper adopts a combination of single-point crossover and multi-point crossover of chromosome fragments of different distribution centers to find a more reasonable way to allocate tasks between distribution centers.

[0062] When performing asexual crossover, first divide the chromosome into multiple segments according to the distribution center, and select any two segments A and B, generate a random number p_c, and if p_c falls within the crossover probability interval, perform a crossover operation. If the number of genes in any segment of chromosome A or B is less than 3, it cannot take two different cut points, and perform a single-point crossover, randomly select a crossover point in A and B, and exchange the segments after the two crossover points; otherwise, perform a multi-point crossover, randomly select two crossover points in A and B, and exchange the segments between the crossover points in the two crossover points.

[0063] The mutation operator is another way to generate new individuals besides crossover. Its connotation is to imitate the mutation in nature and change the value of a gene position on the chromosome with a small probability. This random change increases the diversity of the population, can jump out of the local optimum, and effectively avoid the premature convergence problem of the algorithm. In this paper, the main function of the mutation operator is to optimize the scheduling of a single distribution center and act on the task sequence of a single distribution center. In view of the fact that each natural number in the natural number encoding cannot appear repeatedly, three mutation methods are designed: exchange, flipping and embedding.

[0064] In this chromosome segment, the original delivery tasks of a distribution center are 7, 3, 8, 9, 2. After the positions of nodes 3 and 9 are swapped, the task order becomes 7, 9, 8, 3, 2; after the chromosome segment is flipped, the task order becomes 2, 9, 8, 3, 7; the node 2 at the end of the chromosome is embedded between 3 and 8, and the task order becomes 7, 3, 2, 8, 9. During the algorithm operation, a random number is first generated. If the random number falls within the mutation probability interval, one of the three mutation operations of swapping, flipping, and embedding is performed on the chromosome.

[0065] For further information, see Figure 7,The improved genetic algorithm process is divided into the following steps:

[0066] (1) Parameter setting: Set the population size pop_size, the maximum number of iterations interation, the crossover probability crossover_rate, and the mutation probability mutation_rate.

[0067] (2) Population initialization. Customer nodes are assigned to distribution centers according to their distances, randomly arranged, and connected with breakpoints to produce an initial population pop0 with a chromosome number of population_size. The fitness values ​​of the chromosomes in population pop0 are calculated, and the optimal chromosome in the first generation is found.

[0068] (3) The chromosomes in the previous generation population are randomly selected pop_size-1 times according to the fitness value using the roulette wheel method, and the crossover mutation operation is performed on the selected chromosomes and stored in the new population library.

[0069] (4) Add the best chromosome from the previous generation to the new population library to form a new population new_pop with a capacity of pop_size. Calculate the fitness of the chromosomes in the new population and find the best chromosome.

[0070] (5) Determine whether the algorithm termination condition is reached. In this paper, the termination condition is the maximum number of iterations. If it is reached, stop the calculation and output the result. Otherwise, go to step 3.

[0071] When using a medical material automatic guided vehicle distribution path optimization system and method, compared with the initial random path, the optimized distribution time is shortened by about 35% on average, the cost is reduced by about 25%, and the load variance between AGVs is reduced by about 40%, indicating that the model and solution algorithm have significant effects in improving distribution efficiency and balancing load. Through the construction of the model and the improvement of the algorithm, the distribution path of medical materials is optimized, thereby shortening the distribution time and improving the distribution efficiency.

[0072] What is disclosed above is only one or more preferred embodiments of the present application, and cannot be used to limit the scope of rights of the present application. Ordinary technicians in this field can understand that all or part of the processes of implementing the above embodiments and equivalent changes made according to the claims of the present application are still within the scope covered by the present application.

Claims

1. A method for optimizing the distribution path of medical supplies automatic guided vehicles, characterized in that: The following steps are involved: Conduct research and analysis on hospital layout, understand and analyze personnel and traffic flows, as well as the characteristics and distribution needs of medical supplies; According to the analysis results, select the appropriate AGV type and navigation method; Determine the minimum delivery time and reduce AGV energy consumption and equipment loss costs; Establish a mathematical model, abstract the hospital environment into a directed graph, construct a shortest path model or a multi-objective optimization model, and consider constraints, including time windows and priorities; An improved genetic algorithm is used to solve the model, and the optimal delivery path is found by encoding the AGV's delivery path into chromosomes, designing fitness functions, and performing selection, crossover and mutation operations.

2. The method for optimizing the distribution path of an automated guided vehicle for medical supplies according to claim 1, characterized in that: Conduct research and analysis on hospital layout, understand and analyze personnel and traffic flow, as well as medical supplies characteristics and distribution needs, and the steps also include: The hospital layout includes floor distribution, department locations, passage widths and connections, and a detailed layout diagram.

3. The method for optimizing the distribution path of medical supplies automatic guided vehicles according to claim 2, characterized in that: Conduct research and analysis on hospital layout, understand and analyze personnel and traffic flow, as well as medical supplies characteristics and distribution needs, and the steps also include: Classify medical supplies, analyze the weight, volume, storage conditions, delivery time and priority requirements of each type of supplies, and convert these requirements into constraints for path optimization.

4. The method for optimizing the distribution path of an automated guided vehicle for medical supplies according to claim 1, characterized in that: The improved genetic algorithm is used to solve the model, and the optimal delivery path is found through encoding, fitness function design, selection, crossover and mutation operations. The steps also include: The genetic algorithm includes adopting a path-based encoding method, designing a fitness function, comprehensively considering delivery time, cost and load balance, and converting multi-objectives into single-objective problems in a weighted manner for solution.

5. The method for optimizing the distribution path of an automated guided vehicle for medical supplies as claimed in claim 4, characterized in that: The improved genetic algorithm is used to solve the model, and the optimal delivery path is found through encoding, fitness function design, selection, crossover and mutation operations. The steps also include: Computer simulation software is used to build a virtual scene similar to the actual hospital environment, simulate different medical supplies distribution tasks and various interference factors, and verify the effectiveness and feasibility of the path optimization model and algorithm.

6. A medical supplies automatic guided vehicle distribution path optimization system, characterized in that: A method for optimizing the distribution path of medical supplies by an automated guided vehicle according to any one of claims 1 to 5 is applied.

Images