Base station maintenance path planning method and device and readable storage medium

Abstract

The invention provides a base station maintenance path planning method and device and a readable storage medium, and relates to the technical field of communication, and the method comprises the steps: obtaining a to-be-maintained base station; calculating position information of the to-be-overhauled base stations through a fence map, analyzing the position information of the to-be-overhauled base stations and the environment information by using a clustering algorithm, and classifying the to-be-overhauled base stations; inputting the classification result into a trained first neural network model to obtain an optimal path of the base station to be overhauled; and taking the optimal path as a part of an initial population of a genetic algorithm in a pre-constructed genetic ant colony hybrid algorithm, and further performing path optimization by using the genetic ant colony hybrid algorithm to obtain a final optimal path of the to-be-overhauled base station. According to the method, the device and the readable storage medium, the problems of low inspection efficiency and long path planning time due to the fact that an existing base station maintenance path planning method mainly depends on manual implementation can be solved.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a base station maintenance path planning method, device and readable storage medium. Background Art

[0002] With the development of my country's communications industry, especially the popularization of 5G networks, the number of communication towers has increased year by year, and various operators will build more unmanned communication base stations. As the most critical infrastructure in mobile communication networks, base stations include structural components such as computer rooms, wires, towers, and various complex functional modules, which require regular inspections and maintenance to ensure the stable operation of the equipment. However, most of these base stations are located in remote, sparsely populated, and inconveniently transported areas, which brings many inconveniences to the daily maintenance of base stations. In addition, with changeable weather and frequent natural disasters such as typhoons, towers are easily damaged by harsh environments. These problems can usually only be discovered after the event or during regular inspections. The problems and defects of the existing technology are specifically reflected in the following aspects:

[0003] 1) Base station monitoring is not timely and in place, and is often only discovered when a fault occurs or when inspection personnel conduct inspections, which can easily cause unnecessary losses;

[0004] 2) Currently, base station maintenance mainly relies on manual implementation, and base station maintenance personnel use multiple shifts to conduct inspections. However, manual inspections often plan their itineraries based on subjective judgment, resulting in negligence during the inspection process and low overall inspection efficiency;

[0005] 3) The path planning methods in the prior art have the problems of long planning time, complex design, large amount of calculation and easy to fall into the local optimal solution. Summary of the invention

[0006] The technical problem to be solved by the present invention is to provide a base station maintenance path planning method, device and readable storage medium in response to the above-mentioned shortcomings of the prior art, so as to at least solve the problems that the existing base station maintenance path planning method mainly relies on manual implementation, resulting in low inspection efficiency and long path planning time.

[0007] In a first aspect, the present invention provides a base station maintenance path planning method, the method comprising:

[0008] Get the base station to be repaired;

[0009] Calculate the location information of the base station to be repaired through the fence map, and use the clustering algorithm to analyze the location information and environmental information of the base station to be repaired, and classify the base stations to be repaired;

[0010] Inputting the classification result into the trained first neural network model to obtain the optimal path of the base station to be repaired;

[0011] The optimal path is used as a part of the initial population of the genetic algorithm in the pre-constructed genetic ant colony hybrid algorithm, and the genetic ant colony hybrid algorithm is used to further optimize the path to obtain the final optimal path of the base station to be repaired.

[0012] Furthermore, the obtaining of the base station to be repaired specifically includes:

[0013] Real-time monitoring of each base station is carried out through video surveillance terminals and sensors deployed at each base station side;

[0014] The operating status of each base station is predicted based on the monitored data to obtain the base station to be repaired.

[0015] Furthermore, after calculating the location information of the base station to be repaired through the fence map, the method further includes:

[0016] The fence map is used to present the location information and operating status of each base station in an intuitive manner.

[0017] Furthermore, the genetic ant colony hybrid algorithm includes a genetic algorithm and an ant colony algorithm, and the genetic ant colony hybrid algorithm is used to further perform path optimization to obtain the final optimal path of the base station to be repaired, specifically including:

[0018] The optimal fusion point of the genetic algorithm and the ant colony algorithm is defined as the point when the optimal solution output by the genetic algorithm for n consecutive generations does not change or the genetic algorithm is stagnant;

[0019] According to the number of maintenance personnel and the number and location information of the base stations to be repaired, the genetic algorithm is used to perform preliminary path planning until the best fusion point of the genetic algorithm and the ant colony algorithm is reached to obtain several groups of optimization solutions;

[0020] The plurality of groups of optimization solutions are converted into pheromone distribution initialization parameters of the ant colony algorithm, and the ant colony algorithm is used to further perform path optimization to obtain the final optimal path of the base station to be repaired.

[0021] Furthermore, according to the number of maintenance personnel and the number and location information of the base stations to be repaired, the genetic algorithm is used to perform preliminary path planning until the optimal fusion point of the genetic algorithm and the ant colony algorithm is reached, and several groups of optimization solutions are obtained, specifically including:

[0022] Initializing various parameters of the genetic algorithm according to the number of maintenance personnel and the number and location information of the base stations to be repaired;

[0023] Determine a starting point in the base station to be repaired;

[0024] Setting a fitness function, and encoding the base station to be repaired and the starting point;

[0025] Initialize the population and randomly generate parent chromosomes;

[0026] Partial crossover and mutation;

[0027] Calculate the objective function and fitness value;

[0028] Selection is made according to the roulette method;

[0029] It is determined whether the optimal fusion point is reached. If the optimal fusion point is reached, several groups of optimization solutions are output, otherwise, the steps of executing partial crossover and mutation are returned.

[0030] Furthermore, the ant colony algorithm is used to further optimize the path to obtain the final optimal path of the base station to be repaired, which specifically includes:

[0031] Initialize various parameters of the ant colony algorithm, and place M ants on N base stations, where M≥1 and N≥1;

[0032] Increase the number of iterations by 1;

[0033] According to the state transition probability formula, move to the next base station;

[0034] Move the selected ant to the next base station and record it in the taboo list;

[0035] Determine whether all base stations have been traversed. If so, update the pheromone and calculate the amount of information. Otherwise, return to the step of moving to the next base station according to the state transition probability formula.

[0036] Determine whether the current path is the shortest path. If so, update the path length;

[0037] It is determined whether the number of iterations reaches the maximum number of iterations. If so, the shortest path is output as the final optimal path. Otherwise, the process returns to the step of increasing the number of loops by 1.

[0038] Furthermore, the method further comprises:

[0039] In each iteration of the ant colony algorithm, the volatilization rate of the pheromone is predicted and adjusted using the trained second neural network model according to the current path state and environmental characteristics.

[0040] In a second aspect, the present invention provides a base station maintenance path planning device, the device comprising:

[0041] A module for acquiring base stations to be repaired, used for acquiring base stations to be repaired;

[0042] A base station classification module for maintenance is connected to the base station acquisition module for maintenance, and is used to calculate the location information of the base station for maintenance through a fence map, and use a clustering algorithm to analyze the location information and environmental information of the base station for maintenance, so as to classify the base stations for maintenance;

[0043] An optimal path acquisition module, connected to the base station classification module to be repaired, for inputting the classification result into the trained first neural network model to obtain the optimal path of the base station to be repaired;

[0044] The path re-optimization module is connected to the optimal path acquisition module, and is used to use the optimal path as a part of the initial population of the genetic algorithm in the pre-constructed genetic ant colony hybrid algorithm, and use the genetic ant colony hybrid algorithm to further optimize the path to obtain the final optimal path of the base station to be repaired.

[0045] In a third aspect, the present invention provides a base station maintenance path planning device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to implement the base station maintenance path planning method described in the first aspect above.

[0046] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the base station maintenance path planning method described in the first aspect is implemented.

[0047] The present invention provides a base station maintenance path planning method, device and readable storage medium. First, the base station to be repaired is obtained; then the location information of the base station to be repaired is calculated through the fence map, and the location information and environmental information of the base station to be repaired are analyzed by using a clustering algorithm to classify the base station to be repaired; then the classification result is input into the trained first neural network model to obtain the optimal path of the base station to be repaired; finally, the optimal path is used as a part of the initial population of the genetic algorithm in the pre-constructed genetic ant colony hybrid algorithm, and the genetic ant colony hybrid algorithm is used to further optimize the path to obtain the final optimal path of the base station to be repaired. This application integrates traditional technology with neural network models, fully utilizes the advantages of various algorithms, and further optimizes the path selection process, so as to obtain the optimal path for base station maintenance, shorten the time wasted during the maintenance arrival process, and ensure the efficient operation of the inspection and maintenance work of the base station, which solves the problem that the existing base station maintenance path planning method mainly relies on manual implementation, has low inspection efficiency, and long path planning time. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1This is a flow chart of a base station maintenance path planning method according to Embodiment 1 of the present invention;

[0049] Figure 2 A flow chart of a genetic ant colony hybrid algorithm according to an embodiment of the present invention;

[0050] Figure 3 This is a schematic diagram of the structure of a base station maintenance path planning device according to Embodiment 2 of the present invention;

[0051] Figure 4 This is a structural diagram of a base station maintenance path planning device according to embodiment 3 of the present invention. DETAILED DESCRIPTION

[0052] In order to enable those skilled in the art to better understand the technical solution of the present invention, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0053] It should be understood that the specific embodiments and drawings described herein are only used to explain the present invention rather than to limit the present invention.

[0054] It can be understood that, in the absence of conflict, the various embodiments of the present invention and the various features in the embodiments can be combined with each other.

[0055] It can be understood that, for the convenience of description, the drawings of the present invention only show the parts related to the present invention, while the parts irrelevant to the present invention are not shown in the drawings.

[0056] It can be understood that each unit and module involved in the embodiments of the present invention may correspond to only one physical structure, or may be composed of multiple physical structures, or multiple units and modules may be integrated into one physical structure.

[0057] It can be understood that the terms "first", "second", etc. in the embodiments of the present invention are used to distinguish different objects, or to distinguish different processing of the same object, rather than to describe a specific order of objects.

[0058] It can be understood that, without conflict, the functions and steps marked in the flowcharts and block diagrams of the present invention may occur in an order different from that marked in the drawings.

[0059] It is understood that the flowcharts and block diagrams of the present invention illustrate the possible architectures, functions, and operations of the systems, devices, equipment, and methods according to the various embodiments of the present invention. Each box in the flowchart or block diagram may represent a unit, module, program segment, or code, which contains executable instructions for implementing the specified functions. Moreover, each box or combination of boxes in the block diagram and flowchart may be implemented by a hardware-based system that implements the specified functions, or may be implemented by a combination of hardware and computer instructions.

[0060] It can be understood that the units and modules involved in the embodiments of the present invention can be implemented by software or hardware. For example, the units and modules can be located in a processor.

[0061] Embodiment 1:

[0062] This embodiment provides a base station maintenance path planning method, such as Figure 1 As shown, the method includes:

[0063] Step S101: Acquire a base station to be repaired.

[0064] In this embodiment, the base stations to be repaired are base stations that need to be repaired and are distributed in different locations.

[0065] Optionally, the acquiring the base station to be repaired specifically includes:

[0066] Real-time monitoring of each base station is carried out through video surveillance terminals and sensors deployed at each base station side;

[0067] The operating status of each base station is predicted based on the monitored data to obtain the base station to be repaired.

[0068] In this embodiment, by deploying video surveillance terminals and sensors on the base station side, the operating status of the base station can be monitored at all times, and the monitored data can be transmitted to the background in real time. With the help of advanced technologies such as artificial intelligence, the operating status of the base station can be predicted, the health of the base station can be analyzed, and it can be determined whether each base station is operating normally, thereby obtaining the base station to be repaired.

[0069] Step S102: Calculate the location information of the base station to be repaired through the fence map, and use a clustering algorithm to analyze the location information and environmental information of the base station to be repaired, and classify the base stations to be repaired.

[0070] In this embodiment, the fence map is a map representation method that divides the geographic space into grids, each grid can represent a geographic space unit, and the three-dimensional location information of the base station that needs to be repaired can be calculated through the fence map. At the same time, the clustering algorithm can be applied to path planning. It clusters and groups the distribution of base stations by analyzing obstacles in the environment or the areas where base stations are distributed, thereby simplifying the path planning process. This analysis can specifically use the clustering method of three-dimensional data to combine the location information of the base station with factors such as the terrain and obstacles of the environment, so as to more accurately determine the distribution pattern of the base station and possible influencing factors.

[0071] Optionally, after calculating the location information of the base station to be repaired through the fence map, the method further includes:

[0072] The fence map is used to present the location information and operating status of each base station in an intuitive manner.

[0073] In this embodiment, in order to help maintenance personnel understand the layout and status of the base station more effectively, the fence map can be used to present the location of the base station, equipment status (ie, operating status) and other data in an intuitive manner.

[0074] Step S103: input the classification result into the trained first neural network model to obtain the optimal path of the base station to be repaired.

[0075] In this embodiment, the first neural network algorithm is used to learn the optimal path planning strategy in a given environment. The first neural network can improve its performance in path planning by learning the characteristics and optimal path in the environment. Specifically, the clustered data information (including cluster center points, base stations within clusters, distances between clusters, etc.) is passed as input to the first neural network model, and after processing, a two-dimensional array (i.e., the optimal path) is finally generated as the output result. The two-dimensional array includes information such as the starting point, intermediate nodes, and end point.

[0076] Step S104: taking the optimal path as a part of the initial population of the genetic algorithm in the pre-constructed genetic ant colony hybrid algorithm, and further optimizing the path using the genetic ant colony hybrid algorithm to obtain the final optimal path of the base station to be repaired.

[0077] In this embodiment, by using the optimal path generated by the first neural network as part of the initial population of the genetic algorithm, the efficiency of the search can be improved. At the same time, in the process of research and practice of the prior art, the inventor found that the genetic algorithm searches for the optimal solution of a single individual in the group and uses it as the parent to perform operations such as crossover and mutation to improve the overall superiority of the group. The difference is the ant colony algorithm, which imitates the behavior of ants foraging and uses the pheromones released by ants to find the optimal path. It has strong positive feedback, robustness, fast convergence speed, strong adaptability, good local search ability, and the ability to perform distributed computing. However, in the later stage of operation, the genetic algorithm is prone to information redundancy, resulting in premature algorithm. On the contrary, the ant colony algorithm needs to accumulate pheromones for a long time in the early stage of operation, resulting in a slow search for the optimal path. Therefore, by combining the advantages and disadvantages of the two algorithms, a genetic ant colony hybrid algorithm can be constructed to achieve the complementarity of the algorithms, combine the advantages of the two, optimize path planning, and seek the optimal path for base station maintenance.

[0078] It should be noted that by combining the genetic algorithm with the ant colony algorithm for maintenance route planning, not only can the problem of low solution efficiency caused by insufficient utilization of feedback information by the genetic algorithm be solved, but also the shortcomings of the ant colony algorithm in the early stage of lack of pheromones and slow solution speed can be overcome. By combining the advantages of the two, the most efficient and time-saving maintenance route can be found.

[0079] Optionally, the genetic ant colony hybrid algorithm includes a genetic algorithm and an ant colony algorithm, and further performing path optimization using the genetic ant colony hybrid algorithm to obtain a final optimal path of the base station to be repaired specifically includes:

[0080] The optimal fusion point of the genetic algorithm and the ant colony algorithm is defined as the point when the optimal solution output by the genetic algorithm for n consecutive generations does not change or the genetic algorithm is stagnant;

[0081] According to the number of maintenance personnel and the number and location information of the base stations to be repaired, the genetic algorithm is used to perform preliminary path planning until the best fusion point of the genetic algorithm and the ant colony algorithm is reached to obtain several groups of optimization solutions;

[0082] The plurality of groups of optimization solutions are converted into pheromone distribution initialization parameters of the ant colony algorithm, and the ant colony algorithm is used to further perform path optimization to obtain the final optimal path of the base station to be repaired.

[0083] In this embodiment, the genetic ant colony hybrid algorithm uses a genetic algorithm in the early stage, because the genetic algorithm has a strong evolutionary ability in the early stage of operation, which can make up for the problem of weak search ability of the ant colony algorithm in the early stage of calculation, thereby avoiding the problems of slow operation speed and poor effect. After a certain period of time, as the accumulation of pheromones reaches a certain amount, the search ability and adaptability of the ant colony algorithm are significantly improved. Therefore, the hybrid algorithm can avoid the problem of low search efficiency of the genetic algorithm in the later stage of the algorithm to a certain extent, improve the search efficiency and speed up the output of the optimal path. Therefore, the two algorithms will intersect within a certain period of time, which shows that mixing the two algorithms at the intersection can make up for their respective shortcomings and improve operation efficiency.

[0084] In this embodiment, in order to achieve the best effect of the genetic ant colony hybrid algorithm, it is necessary to determine the important time for the fusion of the two algorithms. The present invention determines whether the algorithm continues to execute the next step or returns to the previous step according to the maximum number of iterations and the fitness value. Among them, the fitness value is used to judge whether the genetic algorithm individuals in the population are close to the optimal solution, or the optimal solution output for n consecutive generations has not changed, which is used as the optimal fusion point of the two algorithms. Specifically, when the optimal solution output by the genetic algorithm for n consecutive generations has not changed, and the ideal solution cannot be output even if it continues to run, it is manifested as the following situation (Formula 2-1); or when the maximum fitness value of the genetic algorithm for n consecutive generations decreases, indicating that the genetic algorithm has degraded, the algorithm has stagnated, and it is difficult to give the optimal solution, it is manifested as the following situation (Formula 2-2). Both of the above situations indicate that the algorithms begin to merge. Therefore, the present invention defines the optimal fusion point of the genetic algorithm and the ant colony algorithm as the time when the optimal solution output by the genetic algorithm for n consecutive generations has not changed or the algorithm has stagnated, so as to ensure that the hybrid algorithm has a strong global search capability.

[0085]

[0086] Where n is the number of iterations, f -n is the average fitness value of n generations, is the maximum fitness value of n generations, With f -n The difference represents the changed fitness value.

[0087] In this embodiment, the algorithm ultimately outputs an optimal path from the starting point to the end point, which is usually represented as a two-dimensional array, where each row represents a node (base station) on the path and each column represents the coordinates of the node. Among them, path allocation is divided into two cases: a) Single-person path: If there is only one maintenance personnel, then the generated path can be directly used to guide the inspection route of the maintenance personnel. The maintenance personnel visit each base station in turn according to the order of the nodes on the path to perform maintenance work. Multi-person path: If there are multiple maintenance personnel, the generated path can be assigned to different maintenance personnel so that they can complete the maintenance tasks of all base stations together.

[0088] Optionally, the genetic algorithm is used to perform preliminary path planning according to the number of maintenance personnel and the number and location information of the base stations to be maintained, until the optimal fusion point of the genetic algorithm and the ant colony algorithm is reached, and several groups of optimization solutions are obtained, specifically including:

[0089] Initializing various parameters of the genetic algorithm according to the number of maintenance personnel and the number and location information of the base stations to be repaired;

[0090] Determine a starting point in the base station to be repaired;

[0091] Setting a fitness function, and encoding the base station to be repaired and the starting point;

[0092] Initialize the population and randomly generate parent chromosomes;

[0093] Partial crossover and mutation;

[0094] Calculate the objective function and fitness value;

[0095] Selection is made according to the roulette method;

[0096] It is determined whether the optimal fusion point is reached. If the optimal fusion point is reached, several groups of optimization solutions are output, otherwise, the steps of executing partial crossover and mutation are returned.

[0097] In this embodiment, various parameters of the genetic algorithm include population size, crossover probability, mutation probability, etc. When studying the maintenance path, the shortest maintenance path is used as the objective function. For example, the objective function can be as follows:

[0098]

[0099] Where minW is the objective function, K represents the set of maintenance personnel, K = {1, 2, ... k}, B represents the number of base stations, is a decision variable, indicating that if the kth maintenance personnel from base station i to base station j performs maintenance, then the value is 1, otherwise the value is 0; D i，j represents the distance from base station i to base station j.

[0100] Optionally, the adopting the ant colony algorithm to further perform path optimization to obtain a final optimal path of the base station to be repaired specifically includes:

[0101] Initialize various parameters of the ant colony algorithm, and place M ants on N base stations, where M≥1 and N≥1;

[0102] Increase the number of iterations by 1;

[0103] According to the state transition probability formula, move to the next base station;

[0104] Move the selected ant to the next base station and record it in the taboo list;

[0105] Determine whether all base stations have been traversed. If so, update the pheromone and calculate the amount of information. Otherwise, return to the step of moving to the next base station according to the state transition probability formula.

[0106] Determine whether the current path is the shortest path. If so, update the path length;

[0107] It is determined whether the number of iterations reaches the maximum number of iterations. If so, the shortest path is output as the final optimal path. Otherwise, the process returns to the step of increasing the number of loops by 1.

[0108] In this embodiment, the state transition probability formula is used to guide the ants to select the next node, and the formula can be as follows.

[0109]

[0110] In the formula, τ i j is the pheromone concentration from base station i to base station j; η i j is the startup information, usually expressed as α and β are the weight coefficients of pheromone and priming information; J k (i) is the set of feasible neighbors of ant k at its current position i.

[0111] Optionally, the method further comprises:

[0112] In each iteration of the ant colony algorithm, the volatilization rate of the pheromone is predicted and adjusted using the trained second neural network model according to the current path state and environmental characteristics.

[0113] In this embodiment, in each iteration of the ant colony algorithm, the trained second neural network model is used to predict and adjust the pheromone volatilization rate according to the current path state and environmental characteristics. The initial value of the pheromone can be set at the beginning of the algorithm and remain unchanged throughout the operation, or adjusted throughout the operation according to a specific strategy.

[0114] Specifically, the second neural network model can dynamically adjust the pheromone volatility rate ρ according to the current path state and environmental characteristics. The training steps of the second neural network model may include:

[0115] 1) Data preparation: Collect historical path planning data, including path status, environmental characteristics, algorithm status, and corresponding parameter settings; and mark the best parameter settings for each set of input features;

[0116] 2) Model selection: Select a suitable neural network model, such as multi-layer perceptron (MLP), convolutional neural network (CNN), etc.

[0117] 3) Model training: Use labeled data to train the neural network model so that it can dynamically adjust the pheromone volatilization rate ρ according to the current path status and environmental characteristics.

[0118] In a specific embodiment, in order to determine the optimal path for base station maintenance, shorten the time lost during the maintenance arrival process, better meet the needs of base station maintenance, and solve the problem of untimely fault maintenance. The base station maintenance path planning method may include the following steps:

[0119] Step 1: Target point distribution analysis: First, analyze the target points and use the fence map to present the base station location, equipment status and other data in an intuitive way to help maintenance personnel understand the layout and status of the base station more effectively.

[0120] Step 2: Intelligently monitor the operation of base stations: collect statistics on base stations that need maintenance and report them to staff in a timely manner.

[0121] Specifically, video surveillance terminals and sensors are installed at each base station to monitor the weather, temperature and its own data of the base station to achieve real-time monitoring of the base station. The monitored data is transmitted to the background, and with the help of advanced technologies such as artificial intelligence, the operating status of the base station is predicted, the health of the base station is analyzed, it is determined whether each base station is operating normally, and the data is fed back in real time.

[0122] Step 3: Base station health monitoring: Based on existing monitoring technology, list the base stations that need maintenance, weight them accordingly, and increase the maintenance priority of these base stations.

[0123] Step 4: Path planning algorithm selection: According to the distribution of target points, select the appropriate maintenance planning algorithm and perform trajectory optimization so that maintenance personnel can cover all target base stations according to the planned path.

[0124] 1. Model Fitting

[0125] It should be noted that the genetic algorithm searches for the optimal solution of a single individual in the group and uses it as the parent to perform crossover, mutation and other operations to improve the overall superiority of the group. The ant colony algorithm is different from the ant colony algorithm, which imitates the foraging behavior of ants and uses the pheromones released by ants to find the optimal path. It has strong positive feedback, robustness, fast convergence speed, strong adaptability, good local search ability, and the ability to perform distributed computing. However, in the later stage of operation, the genetic algorithm is prone to information redundancy, resulting in premature algorithm. On the contrary, the ant colony algorithm needs a long time to accumulate pheromones in the early stage of operation, resulting in a slow search for the optimal path.

[0126] In the process of research and practice of the prior art, the inventors found that: combining the advantages and disadvantages of the two algorithms, a genetic ant colony hybrid algorithm can be constructed to achieve algorithm complementarity. The hybrid algorithm uses a genetic algorithm in the early stage, because the genetic algorithm has a strong evolutionary ability in the early stage of operation, which can make up for the problem of weak search ability of the ant colony algorithm in the early stage of calculation, thereby avoiding the problems of slow operation speed and poor effect. After a certain period of time, as the accumulation of pheromones reaches a certain amount, the search ability and adaptability of the ant colony algorithm are significantly improved. Therefore, the hybrid algorithm can avoid the problem of low search efficiency of the genetic algorithm in the later stage of the algorithm to a certain extent, improve search efficiency and speed up the output of the optimal path. Therefore, the two algorithms will intersect within a certain period of time, which shows that mixing the two algorithms at the intersection can make up for their respective shortcomings and improve operation efficiency.

[0127] In order to achieve the best effect of the hybrid algorithm, it is necessary to determine the important time for the fusion of the two algorithms. The present invention determines whether the algorithm continues to execute the next step or returns to the previous step according to the maximum number of iterations and the fitness value. Among them, the fitness value is used to judge whether the genetic algorithm individual in the population is close to the optimal solution, or the optimal solution output for n consecutive generations has not changed, which is used as the optimal fusion point of the two algorithms. Specifically, when the optimal solution output by the genetic algorithm for n consecutive generations has not changed, and the ideal solution cannot be output even if it continues to run, it is manifested as the following situation (Formula 2-1); or when the maximum fitness value of the genetic algorithm for n consecutive generations decreases, indicating that the genetic algorithm has degraded, the algorithm tends to stagnate, and it is difficult to give the optimal solution, it is manifested as the following situation (Formula 2-2). Both of the above situations indicate that the algorithm begins to merge. Therefore, the present invention defines the optimal fusion point of the genetic algorithm and the ant colony algorithm when the optimal solution output by the genetic algorithm for n consecutive generations has not changed or the algorithm is stagnant, so as to ensure that the hybrid algorithm has a strong global search capability.

[0128]

[0129] Where n is the number of iterations, f -n is the average fitness value of n generations, is the maximum fitness value of n generations, With f -n The difference represents the changed fitness value.

[0130] 2. Related variables and parameter settings

[0131] (1) A: Operation center, initial location

[0132] (2) B: Number of base stations

[0133] (3) N: base station set, N = {0, 1, 2, ... n}

[0134] (4)D i，j : represents the distance from base station i to base station j

[0135] (5) K: maintenance personnel set, K = {1, 2, ... k}

[0136] (6) Decision variable. If the kth maintenance personnel from base station i to base station j performs maintenance, the value is 1, otherwise the value is 0

[0137] (7)B k : The number of base stations that the kth maintenance personnel has maintained this time

[0138] (8)minW: objective function

[0139] 3. Model construction

[0140] (1) Objective function: When studying the maintenance path, the objectives are usually divided into several types, such as shortest distance, lowest cost, and highest timeliness. The present invention takes the shortest maintenance path as the objective function.

[0141] (2) Constraints

[0142] (a) Maintenance personnel constraints

[0143] (b) Base point constraints

[0144]

[0145] Objective function: Formula (2-3) represents the objective function of shortest maintenance distance

[0146] The constraints include:

[0147] 0≤B k ≤B(Formula 2-4)

[0148]

[0149] Formulas 2-4, 2-5, and 2-6 indicate that the number of base stations provided with maintenance services by maintenance personnel shall not exceed the total number of base stations, and each base station can only be served by one group of maintenance personnel.

[0150]

[0151] Formula 2-7 indicates that if the kth maintenance personnel from base station i to base station j provides maintenance service, the value is 1, otherwise the value is 0.

[0152] (3) Setting the fitness function: The size of the fitness function value determines the quality of the individual. It is usually converted into three fitness functions: objective function, objective function maximization, and objective function minimization, as shown in formulas (Formula 2-8), (Formula 2-9), and (Formula 2-10), respectively, where F(x) represents the fitness function, f(x) represents the fitness function value, Cmax represents the maximum fitness function value, and minW represents the objective function with the shortest delivery distance.

[0153]

[0154] F(x)=1 / minW(Formula 2-10)

[0155] (4) State transition formula

[0156] The state transition probability formula is used to guide ants to select the next node, as shown in (Formula 2-11).

[0157]

[0158] Among them, τ i j is the pheromone concentration from base station i to base station j; η i j is the startup information, usually expressed as α and β are the weight coefficients of pheromone and priming information; J k (i) is the set of feasible neighbors of ant k at its current position i.

[0159] 4. Genetic Ant Colony Hybrid Algorithm Process

[0160] Combining the advantages and characteristics of genetic algorithm and ant colony algorithm, the present invention adopts the method of introducing genetic algorithm into the early stage of ant colony algorithm, such as Figure 2 As shown, the specific process is as follows:

[0161] (1) Initialize parameters and determine the number of maintenance personnel and base stations;

[0162] (2) According to the real-time monitoring results of the base stations, confirm whether there are base stations that need maintenance. If so, select the first station for maintenance as the first station for maintenance, and change the initial location and set k;

[0163] It should be noted that the initial point can be dynamically adjusted according to the monitored base station conditions. If there is an abnormal base station, the abnormal base station point is used as the initial station for maintenance, and the path planning is performed for other base stations. If there is no abnormal base station, the path planning is performed according to the first station maintenance location, and the maintenance personnel execute the corresponding maintenance path.

[0164] (3) According to equations (2-1, 2-2, and 2-3), formulate the adaptation function and coding. Encode the base station and the starting point, with 0 representing the initial location and 1-n representing each base station;

[0165] (4) Initialize the population and randomly generate parent chromosomes;

[0166] (5) Partial crossover and mutation;

[0167] (6) Calculate the objective function and fitness value;

[0168] (7) Selection according to the roulette method;

[0169] (8) Determine whether the maximum number of evolution times has been reached. If yes, proceed to step (10); otherwise, proceed to step (9);

[0170] (9) Determine whether the strategy is met according to equations (2-1) and (2-2). If so, proceed to step (10); if not, proceed to step (5);

[0171] (10) outputting several groups of optimization solutions;

[0172] (11) Initialize the parameters of the ant colony algorithm. Convert the optimized solution into pheromone distribution and place M ants on N base stations;

[0173] (12) The number of iterations plus one operation NU = NU + 1;

[0174] (13) According to the state transition probability formula, move to the next base station;

[0175] (14) Move the selected ant M to the next base station, record it in the taboo list, and determine whether all base stations have been traversed. If so, execute step (15), otherwise return to (13);

[0176] (15) Update pheromones and calculate information quantity;

[0177] (16) Determine whether it is the shortest path, if yes, execute step (17), otherwise execute step (18);

[0178] (17) Update path length;

[0179] (18) Determine if NU≥NUmax, and if so, execute step (19); otherwise, execute step (12);

[0180] (19) Terminate the calculation and output the optimal path.

[0181] It should be noted that the algorithm ultimately outputs an optimal path from the starting point to the end point, which is usually represented as a two-dimensional array, where each row represents a node (base station) on the path and each column represents the coordinates of the node. Among them, path allocation is divided into two cases: a) Single-person path: If there is only one maintenance personnel, the generated path can be directly used to guide the inspection route of the maintenance personnel. The maintenance personnel visit each base station in turn according to the order of the nodes on the path to perform maintenance work. Multi-person path: If there are multiple maintenance personnel, the generated path can be assigned to different maintenance personnel so that they can complete the maintenance tasks of all base stations together.

[0182] 5. Specific steps for adjusting parameters using neural network (i.e., second neural network)

[0183] Data preparation: Collect historical path planning data, including path status, environmental characteristics, algorithm status, and corresponding parameter settings; and mark the best parameter settings for each set of input features;

[0184] Model selection: Choose a suitable neural network model, such as multi-layer perceptron (MLP), convolutional neural network (CNN), etc.

[0185] Model training: Use labeled data to train the neural network model so that it can dynamically adjust the pheromone volatilization rate ρ and the pheromone initial value τ0 according to the current path status and environmental characteristics;

[0186] Model application: In each iteration of the ant colony algorithm, the trained neural network model is used to predict and adjust the pheromone volatility rate ρ and the pheromone initial value τ0 according to the current path status and environmental characteristics;

[0187] It should be noted that clustering algorithms can be applied to path planning. It clusters and groups the distribution of base stations by analyzing obstacles in the environment or the areas where base stations are distributed, thereby simplifying the path planning process. The data information processed by clustering is passed as input to the neural network (i.e., the first neural network), and after processing, a two-dimensional array is finally generated as the output result. The maintenance path planning is carried out by combining the genetic algorithm with the ant colony algorithm. This algorithm can not only solve the problem of low solution efficiency caused by insufficient utilization of feedback information by the genetic algorithm, but also overcome the shortcomings of the lack of pheromones and slow solution speed in the early stage of the ant colony algorithm. By combining the advantages of the two, the most efficient and time-saving maintenance route is sought. At the same time, the performance of the genetic ant colony algorithm is optimized by using neural network technology, such as: using neural networks to learn heuristic information and guide the action selection of ants in the ant colony algorithm; using neural networks to adjust the volatility rate and pheromone parameters of the ant colony algorithm, so as to find high-quality paths faster.

[0188] In another specific embodiment, the base station maintenance path planning method is applied to a base station maintenance path planning system, which includes:

[0189] 1. The base station monitoring module collects the current geographical location of the base station, the health status of the base station and the environmental information in the surrounding area, including temperature, humidity, etc., through intelligent monitoring and sensors deployed on the base station, and collects various data of the base station environment in real time;

[0190] 1) Real-time monitoring: Through the camera system, relevant personnel can remotely and timely check the status of the base station;

[0191] 2) Cloud recording: The system can record videos in the cloud to avoid loss of videos due to equipment damage;

[0192] 3) Video sharing: Share videos with relevant personnel through WeChat or APP, or integrate the system into the business system through API to achieve unified management;

[0193] 4) Video playback: When a problem occurs, you can review the recorded video;

[0194] 5) Real-time data upload: The data in the sensor can be uploaded to the server and cloud storage in real time.

[0195] 2. Data management module: the data management module is used to store and manage the collected data;

[0196] 3. The path planning module is used to plan the maintenance path for maintenance personnel and improve the efficiency of maintenance. It takes the location of each base station as the target, and optimizes the trajectory according to the distribution of target points, combining the advantages of genetic algorithm and ant colony algorithm, so that maintenance personnel can cover all target base stations according to the planned path.

[0197] It should be noted that the base station maintenance path planning method provided by the present invention has the following beneficial effects:

[0198] a) Use advanced technology and algorithms to monitor base stations and schedule maintenance, taking full account of the base station's own conditions. By placing appropriate cameras and sensors around the base station, the surrounding environment is monitored in real time, and an alarm is issued in a timely manner if human damage or abnormal movement of objects is found;

[0199] b) Convert the path problem of base station maintenance into a traveling salesman problem including the starting point. Combine the health status of each base station and the distance between base stations to find the optimal path for base station maintenance;

[0200] c) Use clustering algorithms to study and analyze the location of base stations, analyze obstacles in the environment and the distribution areas of base stations, so as to better understand the distribution of base stations and divide the areas. This analysis can use the clustering method of three-dimensional data to combine the location information of base stations with factors such as terrain and obstacles in the environment, so as to more accurately determine the distribution pattern of base stations and possible influencing factors.

[0201] d) Use neural network algorithms to learn the optimal path planning strategy in a given environment. Neural networks can improve their performance in path planning by learning the characteristics and optimal paths in the environment. The algorithm can receive input features of the environment, such as the starting point, target point, obstacle location, etc., and output an optimized two-dimensional array, which includes information such as the starting point, intermediate nodes, and end point. By using the path generated by the neural network as the initial population of the genetic algorithm, the efficiency of the search can be improved.

[0202] e) It integrates genetic algorithm and ant colony algorithm, making full use of the superior performance of genetic algorithm such as parallelism, high efficiency, and global search, combining the advantages of the two, optimizing path planning, and seeking the optimal path for base station maintenance;

[0203] f) A base station path planning system was constructed to ensure efficient inspection and maintenance of base stations.

[0204] The base station maintenance path planning method provided by the embodiment of the present invention first obtains the base station to be repaired; then calculates the location information of the base station to be repaired through the fence map, and uses the clustering algorithm to analyze the location information and environmental information of the base station to be repaired, and classifies the base station to be repaired; then inputs the classification result into the trained first neural network model to obtain the optimal path of the base station to be repaired; finally, the optimal path is used as a part of the initial population of the genetic algorithm in the pre-constructed genetic ant colony hybrid algorithm, and the genetic ant colony hybrid algorithm is used to further optimize the path to obtain the final optimal path of the base station to be repaired. This application integrates traditional technology with neural network models, fully utilizes the advantages of various algorithms, and further optimizes the path selection process, so as to obtain the optimal path for base station maintenance, shorten the time lost during the maintenance arrival process, and ensure the efficient operation of the inspection and maintenance work of the base station, which solves the problem that the existing base station maintenance path planning method mainly relies on manual implementation, has low inspection efficiency, and long path planning time.

[0205] Embodiment 2:

[0206] like Figure 3 As shown, this embodiment provides a base station maintenance path planning device, which is used to execute the above-mentioned base station maintenance path planning method, including:

[0207] A base station to be repaired acquisition module 11 is used to acquire a base station to be repaired;

[0208] The base station classification module 12 is connected to the base station acquisition module 11, and is used to calculate the location information of the base station through the fence map, and use the clustering algorithm to analyze the location information and environmental information of the base station to be repaired, so as to classify the base stations to be repaired;

[0209] The optimal path acquisition module 13 is connected to the base station classification module 12 to be repaired, and is used to input the classification result into the trained first neural network model to obtain the optimal path of the base station to be repaired;

[0210] The path re-optimization module 14 is connected to the optimal path acquisition module 13, and is used to use the optimal path as a part of the initial population of the genetic algorithm in the pre-constructed genetic ant colony hybrid algorithm, and use the genetic ant colony hybrid algorithm to further optimize the path to obtain the final optimal path of the base station to be repaired.

[0211] Optionally, the base station to be repaired acquisition module 11 specifically includes:

[0212] A real-time monitoring unit is used to monitor each base station in real time through video monitoring terminals and sensors deployed on each base station side;

[0213] The operation status prediction unit is used to predict the operation status of each base station according to the monitored data to obtain the base station to be repaired.

[0214] Optionally, the device further comprises:

[0215] The information display module is used to present the location information and operating status of each base station in an intuitive manner using a fence map.

[0216] Optionally, the genetic ant colony hybrid algorithm includes a genetic algorithm and an ant colony algorithm, and the path re-optimization module 14 includes:

[0217] A fusion point definition unit, used to define the optimal fusion point of the genetic algorithm and the ant colony algorithm when the optimal solution output by the genetic algorithm for n consecutive generations does not change or the genetic algorithm is stagnant;

[0218] A first path optimization unit is used to use the genetic algorithm to perform preliminary path planning according to the number of maintenance personnel and the number and location information of the base stations to be repaired, until the optimal fusion point of the genetic algorithm and the ant colony algorithm is reached to obtain several groups of optimization solutions;

[0219] The second path optimization unit is used to convert the plurality of optimization solutions into the pheromone distribution initialization parameters of the ant colony algorithm, and further perform path optimization using the ant colony algorithm to obtain the final optimal path of the base station to be repaired.

[0220] Optionally, the first path optimization unit includes:

[0221] A first initialization unit, used to initialize various parameters of the genetic algorithm according to the number of maintenance personnel and the number and location information of the base stations to be repaired;

[0222] A starting point determination unit, used to determine a starting point in the base station to be repaired;

[0223] A function setting unit, used to set a fitness function and encode the base station to be repaired and the starting point;

[0224] The population initialization unit is used to initialize the population and randomly generate parent chromosomes;

[0225] Crossover mutation unit, used for partial crossover and mutation;

[0226] A first calculation unit, used for calculating the objective function and the fitness value;

[0227] A selection unit for making selections according to the roulette method;

[0228] The first judgment unit is used to judge whether the optimal fusion point is reached. If the optimal fusion point is reached, a plurality of groups of optimization solutions are output; otherwise, the steps of executing partial crossover and mutation are returned.

[0229] Optionally, the second path optimization unit includes:

[0230] A second initialization unit is used to initialize various parameters of the ant colony algorithm and place M ants on N base stations, where M≥1 and N≥1;

[0231] The number increase unit is used to increase the number of iterations by 1;

[0232] A state transfer unit, used to move to the next base station according to a state transfer probability formula;

[0233] A base station transfer unit is used to move the selected ant to the next base station and record it in the taboo list;

[0234] The second judgment unit is used to judge whether all base station sites have been traversed. If they have been traversed, the pheromone is updated and the amount of information is calculated. Otherwise, the step of returning to execute the step of moving to the next base station site according to the state transition probability formula is returned.

[0235] A third judging unit, used to judge whether the current path is the shortest path, and if so, update the path length;

[0236] The fourth judgment unit is used to judge whether the number of iterations reaches the maximum number of iterations. If so, the shortest path is output as the final optimal path. Otherwise, the process returns to the step of increasing the number of loops by 1.

[0237] Optionally, the device further comprises:

[0238] The volatility adjustment module is used to predict and adjust the volatility of the pheromone using the trained second neural network model according to the current path state and environmental characteristics in each iteration of the ant colony algorithm.

[0239] Embodiment 3:

[0240] refer to Figure 4 This embodiment provides a base station maintenance path planning device, including a memory 21 and a processor 22, the memory 21 stores a computer program, and the processor 22 is configured to run the computer program to execute the base station maintenance path planning method in Example 1.

[0241] The memory 21 is connected to the processor 22. The memory 21 may be a flash memory, a read-only memory or other memory. The processor 22 may be a central processing unit or a single-chip microcomputer.

[0242] Embodiment 4:

[0243] This embodiment provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the base station maintenance path planning method in the above-mentioned embodiment 1 is implemented.

[0244] The computer-readable storage medium includes volatile or non-volatile, removable or non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, computer program modules or other data). Computer-readable storage media include, but are not limited to, RAM (Random Access Memory), ROM (Read-Only Memory), EEPROM (Electrically Erasable Programmable read only memory), flash memory or other memory technology, CD-ROM (Compact Disc Read-Only Memory), digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer.

[0245] In summary, the base station maintenance path planning method, device and readable storage medium provided by the embodiment of the present invention first obtain the base station to be repaired; then calculate the location information of the base station to be repaired through the fence map, and use the clustering algorithm to analyze the location information and environmental information of the base station to be repaired, and classify the base station to be repaired; then input the classification result into the trained first neural network model to obtain the optimal path of the base station to be repaired; finally, the optimal path is used as a part of the initial population of the genetic algorithm in the pre-constructed genetic ant colony hybrid algorithm, and the genetic ant colony hybrid algorithm is used to further optimize the path to obtain the final optimal path of the base station to be repaired. This application integrates traditional technology with neural network models, makes full use of the advantages of various algorithms, and further optimizes the path selection process, so as to obtain the optimal path for base station maintenance, shorten the time lost during the maintenance arrival process, ensure the efficient operation of the inspection and maintenance work of the base station, and solves the problem that the existing base station maintenance path planning method mainly relies on manual implementation, has low inspection efficiency, and long path planning time.

[0246] It is to be understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present invention, but the present invention is not limited thereto. For those of ordinary skill in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A base station maintenance path planning method, characterized in that: The method comprises: Get the base station to be repaired; Calculate the location information of the base station to be repaired through the fence map, and use the clustering algorithm to analyze the location information and environmental information of the base station to be repaired, and classify the base stations to be repaired; Inputting the classification result into the trained first neural network model to obtain the optimal path of the base station to be repaired; The optimal path is used as a part of the initial population of the genetic algorithm in the pre-constructed genetic ant colony hybrid algorithm, and the genetic ant colony hybrid algorithm is used to further optimize the path to obtain the final optimal path of the base station to be repaired.

2. The method according to claim 1, characterized in that The obtaining of the base station to be repaired specifically includes: Real-time monitoring of each base station is carried out through video surveillance terminals and sensors deployed at each base station side; The operating status of each base station is predicted based on the monitored data to obtain the base station to be repaired.

3. The method according to claim 1, characterized in that: After calculating the location information of the base station to be repaired through the fence map, the method further includes: The fence map is used to present the location information and operating status of each base station in an intuitive manner.

4. The method according to claim 1, characterized in that: The genetic ant colony hybrid algorithm includes a genetic algorithm and an ant colony algorithm. The genetic ant colony hybrid algorithm is used to further optimize the path to obtain the final optimal path of the base station to be repaired, which specifically includes: The optimal fusion point of the genetic algorithm and the ant colony algorithm is defined as the point when the optimal solution output by the genetic algorithm for n consecutive generations does not change or the genetic algorithm is stagnant; According to the number of maintenance personnel and the number and location information of the base stations to be repaired, the genetic algorithm is used to perform preliminary path planning until the best fusion point of the genetic algorithm and the ant colony algorithm is reached to obtain several groups of optimization solutions; The plurality of groups of optimization solutions are converted into pheromone distribution initialization parameters of the ant colony algorithm, and the ant colony algorithm is used to further perform path optimization to obtain the final optimal path of the base station to be repaired.

5. The method according to claim 4, characterized in that According to the number of maintenance personnel and the number and location information of the base stations to be repaired, the genetic algorithm is used to perform preliminary path planning until the optimal fusion point of the genetic algorithm and the ant colony algorithm is reached, and several groups of optimization solutions are obtained, specifically including: Initializing various parameters of the genetic algorithm according to the number of maintenance personnel and the number and location information of the base stations to be repaired; Determine a starting point in the base station to be repaired; Setting a fitness function, and encoding the base station to be repaired and the starting point; Initialize the population and randomly generate parent chromosomes; Partial crossover and mutation; Calculate the objective function and fitness value; Selection is made according to the roulette method; It is determined whether the optimal fusion point is reached. If the optimal fusion point is reached, several groups of optimization solutions are output, otherwise, the steps of executing partial crossover and mutation are returned.

6. The method according to claim 4, characterized in that The further path optimization using the ant colony algorithm to obtain the final optimal path of the base station to be repaired specifically includes: Initialize various parameters of the ant colony algorithm, and place M ants on N base stations, where M≥1 and N≥1; Increase the number of iterations by 1; According to the state transition probability formula, move to the next base station; Move the selected ant to the next base station and record it in the taboo list; Determine whether all base stations have been traversed. If so, update the pheromone and calculate the amount of information. Otherwise, return to the step of moving to the next base station according to the state transition probability formula. Determine whether the current path is the shortest path. If so, update the path length; It is determined whether the number of iterations reaches the maximum number of iterations. If so, the shortest path is output as the final optimal path. Otherwise, the process returns to the step of increasing the number of loops by 1.

7. The method according to claim 6, characterized in that The method further comprises: In each iteration of the ant colony algorithm, the volatilization rate of the pheromone is predicted and adjusted using the trained second neural network model according to the current path state and environmental characteristics.

8. A base station maintenance path planning device, characterized in that: The device comprises: A module for acquiring base stations to be repaired, used for acquiring base stations to be repaired; A base station classification module for maintenance is connected to the base station acquisition module for maintenance, and is used to calculate the location information of the base station for maintenance through a fence map, and use a clustering algorithm to analyze the location information and environmental information of the base station for maintenance, so as to classify the base stations for maintenance; An optimal path acquisition module, connected to the base station classification module to be repaired, for inputting the classification result into the trained first neural network model to obtain the optimal path of the base station to be repaired; The path re-optimization module is connected to the optimal path acquisition module, and is used to use the optimal path as a part of the initial population of the genetic algorithm in the pre-constructed genetic ant colony hybrid algorithm, and use the genetic ant colony hybrid algorithm to further optimize the path to obtain the final optimal path of the base station to be repaired.

9. A base station maintenance path planning device, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to implement the base station maintenance path planning method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the base station maintenance path planning method according to any one of claims 1 to 7 is implemented.

Images