Intelligent city planning auxiliary decision-making method and system based on brain neural network

Abstract

The invention discloses an intelligent city planning auxiliary decision-making method and system based on a brain neural network, and relates to the technical field of city planning, and the method comprises the following steps: obtaining the multi-dimensional data of a garbage cleaning site of a certain city, inputting the multi-dimensional data of the garbage cleaning site into a preset brain neural network, and carrying out the dimension reduction processing, obtaining low-dimensional data of the garbage cleaning station; constructing a garbage cleaning site graph model based on the garbage cleaning site low-dimensional data; a sliding window mechanism is adopted for the garbage cleaning site graph model, and G garbage key cleaning sites are obtained based on a preset garbage key cleaning site detection algorithm; and based on the G garbage key cleaning sites, performing urban intelligent garbage cleaning auxiliary decision making, and solving the problems that a traditional garbage cleaning site planning method cannot dynamically adapt to a quickly changing garbage generation mode of a city, so that site configuration is unreasonable, and cleaning efficiency is affected.

Description

Technical Field

[0001] The present invention relates to the technical field of urban planning, and more specifically, to an intelligent urban planning auxiliary decision-making method and system based on brain neural network. Background Art

[0002] The planning of garbage cleaning sites is crucial in modern urban management, but it faces many challenges. With the acceleration of urbanization, the urban population continues to increase, and the amount of garbage generated has increased accordingly. Existing cleaning facilities often cannot meet the demand. In addition, high-density residential and commercial areas make the demand for garbage cleaning more urgent, requiring a re-evaluation of the site selection and number. The planning of traditional garbage cleaning sites usually relies on historical experience and intuition for planning, rather than real-time and comprehensive data analysis. This leads to a lack of scientific basis and difficulty in meeting actual needs when dealing with the rapidly changing garbage generation patterns in cities; in addition, the planning methods of traditional garbage cleaning sites are often static and rarely adjusted once set. This method cannot adapt to dynamic changes in garbage generation, residents' needs, and traffic conditions, resulting in unreasonable configuration of garbage cleaning sites and affecting cleaning efficiency. In response to the above problems, the present invention proposes a solution. Summary of the invention

[0003] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present invention provide an intelligent city planning decision-making support method and system based on a neural network. The intelligent city planning decision-making support method and system based on a brain neural network are used to solve the problem that the planning method of traditional garbage cleaning sites cannot dynamically adapt to the rapidly changing garbage generation pattern in the city, resulting in unreasonable site configuration and affecting the cleaning efficiency.

[0004] To achieve the above object, the present invention provides the following technical solutions: The intelligent city planning auxiliary decision-making method based on brain neural network includes the following steps: Acquire multidimensional data of garbage cleaning stations in a certain city, input the multidimensional data of garbage cleaning stations into a preset cerebellar neural network for dimensionality reduction processing, and obtain low-dimensional data of garbage cleaning stations; construct a garbage cleaning station graph model based on the low-dimensional data of garbage cleaning stations; adopt a sliding window mechanism for the garbage cleaning station graph model, and obtain G key garbage cleaning stations based on a preset garbage key cleaning station detection algorithm; and make auxiliary decisions for urban intelligent garbage cleaning based on the G key garbage cleaning stations.

[0005] In a preferred embodiment, the multidimensional data of the garbage cleaning sites are input into a preset cerebellar neural network for dimensionality reduction processing to obtain low-dimensional data of the garbage cleaning sites, specifically: the multidimensional data of the garbage cleaning sites in a certain city are converted into multidimensional vectors of the garbage cleaning sites; the multidimensional vectors of the garbage cleaning sites are discretized using a quantization function to obtain discretized multidimensional vectors of the garbage cleaning sites; based on the preset cerebellar neural network, the discretized multidimensional vectors of the garbage cleaning sites are activated as storage units to obtain k garbage cleaning site storage units; the weights of the k garbage cleaning site storage units are summed to obtain low-dimensional data of the garbage cleaning sites.

[0006] In a preferred embodiment, the discretized garbage cleaning site multidimensional vector is activated as a storage unit based on the preset cerebellar neural network to obtain k garbage cleaning site storage units, specifically: the discretized garbage cleaning site multidimensional vector is combined into the address of the storage unit through a hash function based on the mapping mechanism of the preset cerebellar neural network; the address of the storage unit is activated as k garbage cleaning site storage units based on the activation mechanism of the preset cerebellar neural network.

[0007] In a preferred embodiment, a garbage cleaning site graph model is constructed based on the low-dimensional data of the garbage cleaning sites, specifically: each garbage cleaning site is defined as a node of the garbage cleaning site graph model, and the relationship between two garbage cleaning sites is used as an edge of the garbage cleaning site graph model; the cleaning priority of the garbage cleaning site is calculated as the edge weight of the garbage cleaning site graph model In a preferred embodiment, a sliding window mechanism is adopted for the garbage cleaning site graph model, and G garbage key cleaning sites are obtained based on a preset garbage key cleaning site detection algorithm, specifically: the garbage cleaning site graph model is converted into a garbage cleaning site dynamic graph sequence using a sliding window mechanism to obtain a garbage cleaning site graph model within a sliding time window; based on the garbage cleaning site graph model within the sliding time window, the enhanced Lagrange multiplier formula is used to detect the garbage key cleaning sites to obtain G garbage key cleaning sites.

[0008] In a preferred embodiment, the garbage cleaning site graph model is converted into a garbage cleaning site dynamic graph sequence by using a sliding window mechanism to obtain a garbage cleaning site graph model within a sliding time window, specifically: defining the size T of the sliding window and the sliding step length , and the sliding window is adjusted according to the sliding step size Construct a garbage cleaning site graph model corresponding to each timestamp respectively; update the nodes, edges and edge weights of the garbage cleaning site graph model corresponding to each timestamp to obtain the updated garbage cleaning site graph models corresponding to the M timestamps; convert the updated garbage cleaning site graph models corresponding to the M timestamps into a garbage cleaning site dynamic graph sequence.

[0009] The technical effects and advantages of the neural network-based intelligent city planning decision-making auxiliary method and system of the present invention are as follows: 1. The present invention obtains multidimensional data of garbage cleaning stations in a certain city, inputs the multidimensional data of garbage cleaning stations into a preset cerebellar neural network for dimensionality reduction processing, and obtains low-dimensional data of garbage cleaning stations, which can effectively simplify the data dimension and bring multiple advantages. First, the data of garbage cleaning stations usually have complex multidimensional features, such as geographical location, cleaning frequency, garbage volume, traffic conditions, etc. Through dimensionality reduction processing, the most important key information can be extracted from these high-dimensional features, reducing data redundancy and simplifying the complexity of subsequent analysis and processing. Secondly, the low-dimensional data after dimensionality reduction retains the main structure and pattern of the original data, so that when performing task scheduling and resource allocation, efficient calculations can still be performed based on smaller data sets. At the same time, CMAC, as a fast-learning neural network, can update weights in real time and adapt to dynamically changing data, which is very critical for the ever-changing needs in urban garbage cleaning tasks. Dimensionality reduction can also improve the operating efficiency of the algorithm, reduce storage and computing costs, and thus provide more efficient decision support for path planning and resource scheduling of urban garbage cleaning.

[0010] 2. The present invention constructs a garbage cleaning site graph model, and uses a sliding window mechanism and a preset garbage key cleaning site detection algorithm to obtain G garbage key cleaning sites to achieve dynamic capture and reflection of changes in garbage cleaning needs. This method enables the model to update the relationship between sites based on real-time data, thereby more accurately identifying key garbage cleaning sites. Through the preset key cleaning site detection algorithm, the G most important garbage cleaning sites can be effectively screened out to ensure the rational allocation of resources and cleaning tasks. This dynamic adjustment not only improves the efficiency of garbage cleaning, but also responds to urban changes in a timely manner and reduces resource waste. In addition, the structured approach of the graph model makes the complex garbage cleaning site relationship clearer, providing support for decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 It is a structural schematic diagram of the intelligent city planning auxiliary decision-making method based on neural network of the present invention.

[0012] Figure 2 This is a schematic diagram of the structure of the intelligent city planning auxiliary decision-making system based on neural network of the present invention. DETAILED DESCRIPTION

[0013] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0014] Embodiment 1, Figure 1 The invention provides an intelligent city planning auxiliary decision-making method based on a neural network, comprising the following steps: S1, obtaining multi-dimensional data of garbage cleaning stations in a certain city, inputting the multi-dimensional data of garbage cleaning stations into a preset cerebellar neural network for dimensionality reduction processing, and obtaining low-dimensional data of garbage cleaning stations; In this example, the multidimensional data of garbage cleaning stations in a city is as follows: The multi-dimensional data of a city’s garbage cleaning stations include basic station data, station operation status, station environment and equipment data.

[0015] In this example, the multi-dimensional data of the garbage cleaning station is input into the preset cerebellum neural network for dimensionality reduction processing to obtain the low-dimensional data of the garbage cleaning station, specifically: Convert the multidimensional data of garbage cleaning stations in a city into multidimensional vectors of garbage cleaning stations; The garbage cleaning site multidimensional vector is discretized by using a quantization function to obtain a discretized garbage cleaning site multidimensional vector; Based on the preset cerebellar neural network, the discretized garbage cleaning site multidimensional vector is activated as a storage unit to obtain k garbage cleaning site storage units; The weights of the storage units of the k garbage cleaning sites are summed to obtain the low-dimensional data of the garbage cleaning sites.

[0016] In this example, the garbage cleaning station multidimensional vector is discretized using a quantization function to obtain a discretized garbage cleaning station multidimensional vector, specifically:

[0017] in, is the discretized multidimensional vector of garbage cleaning sites, is the jth feature, for is the jth eigenvalue of garbage cleaning station i, Features quantization step size.

[0018] It should be noted that discretizing the multidimensional vectors of garbage cleaning sites through quantization functions can effectively simplify the complexity of data processing and improve the computational efficiency and robustness of the algorithm. First, discretization converts continuous multidimensional data into finite interval representations, reducing the accuracy requirements of the data, thereby reducing the complexity of the calculation, which is beneficial to subsequent storage and processing. In addition, the discretized vectors are suitable for hash mapping or other index-based storage methods, which facilitates the rapid search and activation of corresponding storage units, especially in applications such as cerebellar neural networks (CMAC). Since discretization can map similar inputs to the same or adjacent discrete values, the system has stronger generalization capabilities and can maintain a stable response to small fluctuations in the input. At the same time, discretization reduces the impact of noise on the system and improves the robustness of data processing. Therefore, through discretization processing, the algorithm can more efficiently and reliably cope with complex multidimensional inputs, especially in large-scale real-time decision-making scenarios.

[0019] In this example, the discretized garbage cleaning site multidimensional vector is activated for storage unit based on the preset cerebellar neural network to obtain k garbage cleaning site storage units, specifically:

[0020] in, The storage unit address generated by the jth hash function, is the weight of the input dimension of the multidimensional vector of the i-th discretized garbage cleaning site in the j-th hash function, is the multidimensional vector of the i-th discretized garbage cleaning site, To represent the modulo operation, The size of the storage unit space.

[0021] It should be noted that the storage unit activation of the discretized garbage cleaning site multidimensional vector based on the preset cerebellar neural network (CMAC) obtains k garbage cleaning site storage units, which has many advantages. First, by activating multiple storage units, CMAC can efficiently process complex multidimensional input data, and the network capacity is large and can store more information. Secondly, activating k storage units helps to improve the generalization ability of the system, because CMAC is essentially a local learning method, and the activated storage units cover multiple neighboring areas of the input vector, which can better handle similar but not completely identical inputs, thereby enhancing the robustness of the system. In addition, since the search and learning process of CMAC is based on hash mapping to storage units, the computational overhead is low and it is suitable for real-time decision-making scenarios. The multiple activated storage units provide redundancy, which helps to reduce overfitting and improve the adaptability and accuracy of the algorithm in a dynamic environment. Therefore, this activation mechanism can improve the generalization performance and ability to process multidimensional data of the system while ensuring computational efficiency.

[0022] In this example, the weights of the k garbage cleaning site storage units are summed to obtain the low-dimensional data of the garbage cleaning site, specifically:

[0023] in, Low-dimensional data for garbage cleaning sites, is the weight value of each garbage cleaning station storage unit, The number of storage units for activated waste cleaning stations.

[0024] It should be noted that the discretized garbage cleaning station multidimensional vector is activated by the preset hash function, and each storage unit has a corresponding weight value. The original multidimensional data of the garbage cleaning station may contain a lot of complex information, such as geographical location, garbage volume, cleaning frequency, traffic conditions, etc. These data dimensions are very high, and directly processing high-dimensional data will increase the computational complexity and storage requirements. Through dimensionality reduction processing, low-dimensional data can simplify these complex information, making it easier to process and analyze, thereby improving the efficiency and performance of the algorithm.

[0025] S2, constructs a garbage cleaning site graph model based on the low-dimensional data of garbage cleaning sites; In this example, a garbage cleaning site graph model is constructed based on the low-dimensional data of garbage cleaning sites, specifically: Each garbage cleaning site is defined as a node of the garbage cleaning site graph model, and the relationship between two garbage cleaning sites is used as an edge of the garbage cleaning site graph model; The cleaning priorities of the garbage cleaning sites are calculated as the edge weights of the garbage cleaning site graph model.

[0026] In this example, each garbage cleaning site is defined as a node of the garbage cleaning site graph model, and the relationship between two garbage cleaning sites is used as an edge of the garbage cleaning site graph model, specifically: Assume that the garbage cleaning site graph model G=(V,E), where V is a node set and each node Corresponding to a garbage cleaning station , assuming that there are n garbage cleaning sites in a city, there are n nodes in the garbage cleaning site graph model: ; Two garbage washing stations and The relationship between Bian said, It means that the two sites are related in some dimension (such as geographical proximity, related cleaning frequency, changes in garbage volume, etc.), and their strength can be measured by weight.

[0027] It should be noted that defining each garbage cleaning site as a node of the garbage cleaning site graph model and taking the relationship between two garbage cleaning sites as the edge of the garbage cleaning site graph model has many advantages. First, this graph model can intuitively display the complex relationship between sites. By using nodes to represent sites and edges to represent the association between sites, such as geographical proximity and garbage volume correlation, it simplifies the representation of data and makes the problem structured. Secondly, by assigning weights to each edge, the relative importance or influence between sites can be quantified, so that when scheduling optimization and resource allocation are carried out, the focus can be focused on key sites and their related connection relationships. In addition, mature graph algorithms such as shortest path and centrality analysis can be directly applied based on the graph structure, which helps to find bottlenecks or optimization directions in the system. By constructing this model, the urban garbage cleaning system can more efficiently plan cleaning routes, reasonably allocate resources, and improve the overall operating efficiency by analyzing key sites, ensuring the stability and sustainability of the system.

[0028] S3, adopting a sliding window mechanism for the garbage cleaning site graph model, and obtaining G garbage key cleaning sites based on a preset garbage key cleaning site detection algorithm; In this example, a sliding window mechanism is used for the garbage cleaning site graph model, and G garbage key cleaning sites are obtained based on a preset garbage key cleaning site detection algorithm, specifically: The garbage cleaning site graph model is converted into a garbage cleaning site dynamic graph sequence using a sliding window mechanism, and the garbage cleaning site graph model within the sliding time window is obtained; Based on the garbage cleaning site graph model within the sliding time window, the enhanced Lagrange multiplier formula is used to detect the key garbage cleaning sites, and G key garbage cleaning sites are obtained.

[0029] In this example, a sliding window mechanism is used to transform the garbage cleaning site graph model into a garbage cleaning site dynamic graph sequence, and the garbage cleaning site graph model within the sliding time window is obtained, specifically: Define the size T of the sliding window and the sliding step size , and the sliding window is adjusted according to the sliding step size Construct the garbage cleaning site graph model corresponding to each timestamp respectively; The nodes, edges, and edge weights of the garbage cleaning site graph model corresponding to each timestamp are updated to obtain updated garbage cleaning site graph models corresponding to M timestamps; The garbage cleaning site graph model corresponding to the updated M timestamps is converted into a garbage cleaning site dynamic graph sequence.

[0030] It should be noted that the sliding window mechanism is used to process the garbage cleaning site graph model, and GGG garbage key cleaning sites are obtained based on the preset key cleaning site detection algorithm, which has multiple advantages. First, the sliding window mechanism can dynamically capture the changes of garbage cleaning sites in different time periods, ensuring that the algorithm reflects the update of site status in real time, thereby realizing flexible site management and cleaning task planning. Second, the key site detection algorithm can identify the sites with the greatest impact in the garbage cleaning network. These sites are often the nodes with the highest cleaning demand or the most significant impact on the operation of the entire system. By focusing on these key sites, the system can prioritize resource allocation, reduce overall cleaning time, and improve operational efficiency. In addition, combined with the analysis of the time dimension, it can help discover periodic fluctuations or emergencies in cleaning demand, and enhance the prediction ability and response speed of garbage cleaning demand. Overall, this method can optimize resource allocation, reduce scheduling complexity, and ensure that the key nodes of the garbage cleaning network are cleaned in time at any time, thereby improving the robustness and efficiency of the system.

[0031] In this example, the Lagrange multiplier formula is enhanced as follows:

[0032] in, is the impact weight of each garbage cleaning station, is a binary variable of node i, indicating whether the node is a key site (1 indicates a key site, 0 indicates a non-key site), is the Laplace matrix of the garbage cleaning site graph model, is the overall connectivity of the garbage cleaning site graph model, Number of critical cleaning stations for garbage.

[0033] S4, based on G key garbage cleaning stations, makes auxiliary decisions on urban intelligent garbage cleaning.

[0034] In this example, based on G key garbage cleaning stations, urban intelligent garbage cleaning auxiliary decision-making is carried out, specifically: Priority task scheduling: Based on the G detected critical garbage cleaning sites, garbage cleaning vehicles are first sent to these critical garbage cleaning sites for processing to avoid environmental or traffic problems caused by garbage accumulation. Non-critical garbage cleaning sites can be flexibly arranged according to changes in cleaning needs and remaining resources; Dynamic resource allocation: Based on the G detected key garbage cleaning sites, cleaning resources (such as cleaning vehicles and personnel) are allocated to the key sites first. For example, if a key site is frequently detected as having heavy garbage accumulation, the cleaning frequency can be increased or more vehicles can be deployed.

[0035] Embodiment 2, Figure 2 The present invention provides an intelligent city planning decision-making support system based on a brain neural network, including a data acquisition module, a model building module, a garbage key cleaning site detection module, and a garbage cleaning decision module: The data acquisition module acquires the multi-dimensional data of garbage cleaning stations in a certain city, inputs the multi-dimensional data of garbage cleaning stations into a preset cerebellar neural network for dimensionality reduction processing, and obtains low-dimensional data of garbage cleaning stations; A model building module builds a garbage cleaning site graph model based on the low-dimensional data of garbage cleaning sites, and calculates the cleaning priority of the garbage cleaning site as the weight of the edge in the garbage cleaning site graph model; The garbage key cleaning site detection module adopts a sliding window mechanism for the garbage cleaning site graph model and obtains G garbage key cleaning sites based on a preset garbage key cleaning site detection algorithm; The garbage cleaning decision module is used to make auxiliary decisions on urban intelligent garbage cleaning based on G key garbage cleaning sites.

[0036] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0037] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0038] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0039] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0040] The above are only specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0041] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A brain neural network-based intelligent city planning decision-making assistance method, characterized in that: The steps include: Acquire multidimensional data of garbage cleaning stations in a city, input the multidimensional data of the garbage cleaning stations into a preset cerebellar neural network for dimensionality reduction processing, and obtain low-dimensional data of the garbage cleaning stations; Constructing a garbage cleaning site graph model based on the low-dimensional data of the garbage cleaning sites; A sliding window mechanism is adopted for the garbage cleaning site graph model, and G garbage key cleaning sites are obtained based on a preset garbage key cleaning site detection algorithm; Based on the G key garbage cleaning sites, auxiliary decision-making for urban intelligent garbage cleaning is carried out.

2. The method for assisting decision-making in intelligent city planning based on brain neural network according to claim 1 is characterized in that: The multi-dimensional data of the garbage cleaning site is input into a preset cerebellar neural network for dimensionality reduction processing to obtain low-dimensional data of the garbage cleaning site, specifically: Converting the multidimensional data of garbage cleaning stations in the city into a multidimensional vector of garbage cleaning stations; The garbage cleaning site multidimensional vector is discretized using a quantization function to obtain a discretized garbage cleaning site multidimensional vector; Based on a preset cerebellar neural network, the discretized garbage cleaning site multidimensional vector is activated as a storage unit to obtain k garbage cleaning site storage units; The weights of the k garbage cleaning site storage units are summed to obtain low-dimensional data of the garbage cleaning sites.

3. The intelligent city planning decision-making assistance method based on brain neural network according to claim 2 is characterized in that: The preset cerebellar neural network activates the storage unit of the discretized garbage cleaning site multidimensional vector to obtain k garbage cleaning site storage units, specifically: Based on the mapping mechanism of the preset cerebellar neural network, the discretized garbage cleaning site multi-dimensional vector is combined into the address of the storage unit through a hash function; Based on the activation mechanism of the preset cerebellar neural network, the address of the storage unit is activated as k garbage cleaning site storage units.

4. The method for assisting decision-making in intelligent city planning based on brain neural network according to claim 1 is characterized in that: A garbage cleaning site graph model is constructed based on the low-dimensional data of the garbage cleaning sites, specifically: Each garbage cleaning site is defined as a node of the garbage cleaning site graph model, and the relationship between two garbage cleaning sites is used as an edge of the garbage cleaning site graph model; The cleaning priorities of the garbage cleaning sites are calculated as edge weights of the garbage cleaning site graph model.

5. The intelligent city planning decision-making assistance method based on brain neural network according to claim 1 is characterized in that: The garbage cleaning site graph model adopts a sliding window mechanism, and obtains G garbage key cleaning sites based on a preset garbage key cleaning site detection algorithm, specifically: The garbage cleaning site graph model is converted into a garbage cleaning site dynamic graph sequence by using a sliding window mechanism to obtain a garbage cleaning site graph model within a sliding time window; Based on the garbage cleaning site graph model within the sliding time window, the key garbage cleaning sites are detected by using an enhanced Lagrange multiplier formula to obtain G key garbage cleaning sites.

6. The method for assisting decision-making in intelligent city planning based on brain neural network according to claim 5 is characterized in that: The sliding window mechanism is used to convert the garbage cleaning site graph model into a garbage cleaning site dynamic graph sequence to obtain the garbage cleaning site graph model within the sliding time window, specifically: Define the size T of the sliding window and the sliding step size , and the sliding window is adjusted according to the sliding step size Construct the garbage cleaning site graph model corresponding to each timestamp respectively; The nodes, edges and edge weights of the garbage cleaning site graph model corresponding to each timestamp are updated to obtain updated garbage cleaning site graph models corresponding to the M timestamps; The garbage cleaning site graph models corresponding to the updated M timestamps are converted into a garbage cleaning site dynamic graph sequence.

7. The method for assisting decision-making in intelligent city planning based on brain neural network according to claim 3 is characterized in that: The garbage cleaning site multidimensional vector is discretized using a quantization function to obtain a discretized garbage cleaning site multidimensional vector, specifically: in, is the discretized multidimensional vector of garbage cleaning sites, is the jth feature, is the jth eigenvalue of garbage cleaning station i, Features The quantization step size, min() is the minimum function.

8. The method for assisting decision-making in intelligent city planning based on brain neural network according to claim 3 is characterized in that: The mapping mechanism based on the preset cerebellar neural network combines the discretized garbage cleaning site multidimensional vectors into the address of the storage unit through a hash function, specifically: in, The storage unit address generated by the jth hash function, is the weight of the input dimension of the multidimensional vector of the i-th discretized garbage cleaning site in the j-th hash function, is the multidimensional vector of the i-th discretized garbage cleaning site, To represent the modulo operation, The size of the storage unit space.

9. The method for assisting decision-making in intelligent city planning based on brain neural network according to claim 5 is characterized in that: The enhanced Lagrange multiplier formula is specifically: in, is the impact weight of each garbage cleaning station, is a binary decision variable of node i, indicating whether the node is a key site (1 indicates a key site, 0 indicates a non-key site), is the Laplace matrix of the garbage cleaning site graph model, is the overall connectivity of the garbage cleaning site graph model, Number of critical cleaning stations for garbage.

10. A system using the brain neural network-based intelligent city planning decision-making support method as described in any one of claims 1 to 9, characterized in that: It includes data acquisition module, model building module, garbage key cleaning site detection module, and garbage cleaning decision module: A data acquisition module is used to acquire multi-dimensional data of garbage cleaning stations in a certain city, and input the multi-dimensional data of the garbage cleaning stations into a preset cerebellar neural network for dimensionality reduction processing to obtain low-dimensional data of the garbage cleaning stations; A model building module, used to build a garbage cleaning site graph model based on the low-dimensional data of the garbage cleaning sites, and calculate the cleaning priority of the garbage cleaning sites as the weight of the edges in the garbage cleaning site graph model; A garbage key cleaning site detection module, used to adopt a sliding window mechanism for the garbage cleaning site graph model, and obtain G garbage key cleaning sites based on a preset garbage key cleaning site detection algorithm; The garbage cleaning decision module is used to make auxiliary decisions on urban intelligent garbage cleaning based on the G key garbage cleaning sites.

Images