Charging demand intelligent prediction method and system based on video analysis

Abstract

The invention discloses an intelligent charging demand prediction method and system based on video analysis, and relates to the technical field of computer platform load balancing, and the method comprises the steps: collecting first feature data of a first object, carrying out the preprocessing of the first feature data, analyzing the first feature data, and obtaining influence data; predicting a future first demand trend in combination with the historical data and the first demand, formulating an early warning mechanism, carrying out scheduling optimization on the system through the future first demand trend, and carrying out first adjustment on a prediction scheduling link according to a real-time feedback mechanism. According to the charging demand intelligent prediction method based on video analysis, the improved YOLOv5 algorithm is used for vehicle detection and identification, and the vehicle type and state are accurately identified by optimizing the feature extraction network and the loss function.

Description

Technical Field

[0001] The present invention relates to the technical fields of computer vision and deep learning, and specifically to an intelligent prediction method and system for charging demand based on video analysis. Background Art

[0002] With the rapid development of new energy vehicles, the planning, construction, operation and management of charging infrastructure face huge challenges. Existing charging demand prediction methods mainly rely on historical charging data and user questionnaires, suffering from problems such as lagging data acquisition, limited sample size, and low prediction accuracy. In addition, traditional methods are difficult to capture the dynamic change characteristics of charging demand and cannot achieve real-time early warning and intelligent scheduling.

[0003] Aiming at the deficiencies of existing charging demand prediction methods, this patent proposes an intelligent prediction system and method for charging demand based on video analysis. Through real-time collection and analysis of video surveillance data, combined with improved deep learning algorithms, accurate identification and classification of vehicles in the charging area are achieved, and a mapping relationship between vehicle models and charging demand is established, thereby improving the accuracy and real-time performance of charging demand prediction.

[0004] Based on computer vision and deep learning technologies, this patent designs a complete charging demand prediction system. First, vehicle image data is collected through a video surveillance system, and an improved YOLOv5 algorithm is used for vehicle detection and recognition; secondly, a vehicle model - charging demand mapping model is established to extract vehicle feature information; then, a prediction algorithm is designed based on spatio-temporal sequence data to achieve accurate prediction of charging demand; finally, the prediction results are displayed through a visualization interface and early warning analysis is carried out. Summary of the Invention

[0005] In view of the above existing problems, the present invention is proposed.

[0006] Therefore, the technical problem solved by the present invention is: existing electric vehicle charging demand prediction methods have deficiencies in accuracy, real-time performance, and resource allocation efficiency, and the problem of realizing automated and accurate charging demand prediction by combining real-time video surveillance data with deep learning technologies.

[0007] To solve the above technical problems, the present invention provides the following technical solution: an intelligent prediction method for charging demand based on video analysis, including collecting first feature data of a first object and preprocessing the first feature data;

[0008] Analyzing the first feature data to obtain influencing data, combining historical data with the influencing data to predict the future first demand trend and formulating an early warning mechanism;

[0009] Optimizing the system through the future first demand trend and making a first adjustment to the prediction and scheduling link according to the real-time feedback mechanism.

[0010] As a preferred solution of the intelligent prediction method for charging demand based on video analysis according to the present invention, wherein: the first feature data includes parameter data during the operation of the first object.

[0011] As a preferred solution of the intelligent prediction method for charging demand based on video analysis according to the present invention, wherein: the influencing data includes data that has a direct impact on the overall operation effect during the operation of the first object.

[0012] The first demand includes the values that the parameter data of the first object needs to reach during operation.

[0013] As a preferred solution of the intelligent prediction method for charging demand based on video analysis according to the present invention, wherein: the scheduling optimization includes allocating and adjusting the resource utilization efficiency according to the future first demand trend.

[0014] As a preferred solution of the intelligent prediction method for charging demand based on video analysis according to the present invention, wherein: the first adjustment includes adjusting the accuracy of the prediction model according to the actual demand obtained by the real-time feedback mechanism.

[0015] As a preferred solution of the intelligent prediction method for charging demand based on video analysis according to the present invention, wherein: the process of analyzing the first feature data to obtain the influencing data includes optimizing the YOLOv5 algorithm, using the optimized YOLOv5 algorithm to analyze the parameter data during the operation of the first object, and detecting and identifying the vehicle models therein.

[0016] The optimization of the YOLOv5 algorithm includes calculating the contribution of the weights and bias terms of each convolutional kernel to the model error through the backpropagation algorithm, and updating the weights according to this information to achieve the specific optimization of the YOLOv5 algorithm.

[0017] The prediction process includes establishing a prediction model using a dual attention mechanism, capturing the dynamic changes of temporal features and extracting key information in the spatial dimension, and generating the final prediction result through a non-linear activation function.

[0018] The early warning mechanism includes dynamically setting the warning threshold according to the mean, standard deviation and adjustment parameters of historical data to set an adaptive early warning mechanism.

[0019] As a preferred solution of the intelligent prediction method for charging demand based on video analysis according to the present invention, wherein: the process of allocation and adjustment includes calculating the charging pile efficiency index and allocating resources according to the efficiency index.

[0020] The adjustment of the accuracy of the prediction model includes solving the optimal system parameter configuration through a multi-objective optimization algorithm.

[0021] Another object of the present invention is to provide an intelligent prediction system for charging demand based on video analysis, which can solve the problems of high error rate and inability to adapt to dynamic changing environments in current charging demand prediction methods through real-time video analysis and multi-modal data integration.

[0022] As a preferred solution of the intelligent prediction system for charging demand based on video analysis according to the present invention, it includes: a collection and processing module, an analysis and prediction module, and an optimization and adjustment module.

[0023] The collection and processing module is used to collect the first feature data of the first object and preprocess the first feature data.

[0024] The analysis and prediction module is used to analyze the first feature data to obtain influence data, and combine historical data with the influence data to predict the future first demand trend and formulate an early warning mechanism.

[0025] The optimization and adjustment module is used to schedule and optimize the system according to the future first demand trend, and make a first adjustment to the prediction and scheduling link according to the real-time feedback mechanism.

[0026] A computer device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of the intelligent prediction method for charging demand based on video analysis are realized.

[0027] A computer-readable storage medium stores a computer program thereon. When the computer program is executed by a processor, the steps of the intelligent prediction method for charging demand based on video analysis are realized.

[0028] The beneficial effects of the present invention: The intelligent prediction method for charging demand based on video analysis provided by the present invention uses an improved YOLOv5 algorithm for vehicle detection and recognition. By optimizing the feature extraction network and loss function, the vehicle type and status are accurately identified. A probability model for charging demand is established using a mapping model based on vehicle characteristics and a Bayesian probability model, enhancing the accuracy and reliability of the prediction. Through real-time video data processing technology, combined with a high-speed data stream processing framework to receive and process video data in real time, the prediction results are quickly generated, greatly reducing the overall response time from data collection to decision output. Through a spatio-temporal sequence prediction model with a dual attention mechanism, considering both the time and space dimensions, the charging demand at different times and locations is predicted. The allocation of charging piles and power supply is dynamically adjusted in combination with the prediction results to optimize the overall resource allocation. The early warning is dynamically adjusted according to the prediction and actual usage conditions to improve the management efficiency. The present invention achieves better results in terms of prediction accuracy, response time, and optimization of resource allocation. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the accompanying drawings required for the description of the embodiments. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can be obtained based on these drawings.

[0030] Figure 1 It is the overall flowchart of an intelligent prediction method for charging demand based on video analysis provided by the first embodiment of the present invention.

[0031] Figure 2 It is the system flowchart of an intelligent prediction method for charging demand based on video analysis provided by the first embodiment of the present invention. Specific Embodiments

[0032] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will make a detailed description of the specific embodiments of the present invention in conjunction with the accompanying drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0033] Embodiment 1, referring to Figure 1-2 , which is an embodiment of the present invention, provides an intelligent prediction method for charging demand based on video analysis, including:

[0034] S1: Collect the first feature data of the first object and preprocess the first feature data.

[0035] Collect real-time data of the charging area and perform image enhancement and noise removal on the real-time data.

[0036] In this embodiment, the first object is specifically the charging area, and the first feature data is specifically the real-time data collected by a high-definition video monitoring device for the real-time data of the charging area. The specific formula is as follows.

[0037] I out (x,y) = α·I in (x,y) + β

[0038] Among them, I in (x,y) and I out (x,y) represent the input and output images, and α and β represent the enhancement functions.

[0039] Image enhancement is performed by the adaptive histogram equalization method, and the enhancement parameters are dynamically adjusted according to the scene illumination conditions to improve the image quality. The improved Gaussian filtering algorithm is used to remove noise. By adaptively adjusting the filtering parameters, the noise interference is removed while maintaining the image edge features. The specific formula is as follows.

[0040]

[0041] Among them, σ represents the standard deviation of the Gaussian function, and (x, y) represents the pixel coordinates.

[0042] In an alternative embodiment, the first object can also be an urban road network, and the first feature data is real-time road video data. The real-time road video data is collected by traffic cameras, and the adaptive histogram equalization is used to enhance the video image quality to ensure that the vehicle flow is clearly visible in case of insufficient lighting. The improved Gaussian filtering algorithm is used to remove the noise in the image, highlighting the vehicle flow edge, and generating a clear traffic video for subsequent vehicle detection and vehicle flow density analysis.

[0043] S2: Analyze the first feature data to obtain the influencing data, combine the historical data with the first demand to predict the future first demand trend and formulate an early warning mechanism.

[0044] Identify and analyze the preprocessed real-time data of the charging area to obtain vehicle information, establish a mapping model between vehicle information and charging demand to predict the charging demand of vehicles, establish a spatio-temporal sequence prediction model to predict the subsequent charging demand of vehicles, and set early warning indicators according to historical data and subsequent charging demand.

[0045] In this embodiment, the first demand is specifically the charging demand. The improved YOLOv5 algorithm is used to detect and identify vehicles, and the weight coefficients and bias terms are adjusted to extract the feature map to improve the detection accuracy. The specific formula is as follows.

[0046]

[0047] Among them, F i represents the i-th layer feature map, w i represents the weight coefficient, and b represents the bias term.

[0048] The loss function is improved by combining the weighted classification loss, box position loss, and object confidence loss to comprehensively evaluate the model performance. The specific formula is as follows.

[0049] L total = λ1L cls + λ2L bax + λ3L obj

[0050] Among them, L cls 、Lbax , L obj are respectively represented as the classification loss, the bounding box location loss, and the object confidence loss, and λ1 is represented as the weight coefficient.

[0051] A mapping model is established based on the obtained vehicle models and charging requirements to predict the charging requirements of vehicle models. The charging requirement model of the mapping relationship is as follows.

[0052] D i = f(V i , B i , T i )

[0053] Among them, D i represents the charging requirement of the i-th type of vehicle model, V i represents the vehicle feature vector, B i represents the battery capacity, and T i represents the charging duration.

[0054] The spatio-temporal sequence prediction model calculates the relationship between the hidden state at the current moment and the previous moment through time attention. The specific formula is as follows.

[0055] α t = softmax(W t [h t , s t-1 )

[0056] Among them, h t represents the hidden state at the current moment, s t-1 represents the state at the previous moment, and W t represents the weight matrix.

[0057] The importance of states at different moments is adjusted through the weight matrix. The spatial attention calculation focuses on the relationship between the spatial features and the temporal context vector. The specific formula is as follows.

[0058] β s = softmax(W s [x s , c t )

[0059] Among them, x s represents the spatial feature, and c t represents the temporal context vector.

[0060] The prediction output calculation combines the results of time attention and spatial attention, and generates the final prediction output through a non-linear activation function and a feature concatenation operation to obtain the temporal and spatial changes of the predicted charging requirements. The specific formula is as follows.

[0061] y t= g([α t ·h t ; β s ·x s )

[0062] where g represents a non - linear activation function, and [α t ·h t ; β s ·x s represents a feature concatenation operation.

[0063] The warning index is obtained through weighted calculation by calculating the load, time, and space - dimension weight coefficients. The warning threshold is dynamically set according to the mean, standard deviation, and adjustment parameter of the historical warning index. The specific warning analysis formula is as follows.

[0064] R = w1R load + w2R time + w3R space

[0065] where R load , R time , R space represent the warning indexes of the load, time, and space dimensions respectively, w i represents the weight coefficient, and R represents the warning index.

[0066]

[0067] where represents the mean of the historical warning index, std(R) represents the standard deviation, and μ, σ represent the adjustment parameters.

[0068] In an alternative embodiment, the first demand can also be the traffic flow demand. The improved YOLOv5 algorithm is used to detect and identify vehicles in the road video, the weight coefficients and bias terms are adjusted to extract the feature map, and the accuracy of detection is improved. The weighted combination of the classification loss, box - position loss, and object confidence loss is used to comprehensively evaluate the model performance. A mapping model between vehicle information and traffic flow demand is established based on the detected vehicle information to predict the road traffic flow demand. The relationship between the hidden states at the current moment and the previous moment is calculated through the time attention mechanism, and combining time attention and space attention, the final traffic flow demand prediction result is generated through a non - linear activation function and a feature concatenation operation. The warning index is obtained through weighted calculation by the load, time, and space - dimension weight coefficients, and the warning threshold is dynamically set according to the mean, standard deviation, and adjustment parameter of the historical warning index.

[0069] S3: Optimize the system through the future first - demand trend, and make the first adjustment to the prediction and scheduling link according to the real - time feedback mechanism.

[0070] Adjust the allocation of resource utilization efficiency according to the future charging demand trend, obtain the actual demand based on the real-time feedback mechanism, and adjust the accuracy of the prediction model.

[0071] In this embodiment, the first adjustment is specifically to adjust the allocation of resources. In the process of allocation adjustment, first evaluate the performance indicators of the accuracy rate, root mean square error, and mean absolute percentage error of the prediction formula. The specific evaluation formulas are as follows.

[0072]

[0073]

[0074] Among them, ACC represents the predicted accurate value, RMSE represents the root mean square error, and MAPE represents the mean absolute percentage error.

[0075] Formulate a real-time scheduling strategy according to the prediction result. The scheduling decision considers the scheduling strategy at time t and the utility function under the given charging demand. Improve the efficiency index of the charging pile at time t to achieve the optimal allocation and utilization of resources. The specific scheduling decision formula is as follows.

[0076] S t = argmax s U(s|D t )

[0077] Among them, S t represents the scheduling strategy at time t, and U(s|D t ) represents the utility function under the given charging demand.

[0078]

[0079] Among them, E i (t) represents the efficiency index of the i-th charging pile at time t.

[0080] Adopt a multi-objective optimization method to improve the performance. The optimization objectives include prediction accuracy and response time. Consider the constraint conditions to ensure the feasibility and effectiveness of the optimization process. The specific optimization algorithm formula and constraint conditions are as follows.

[0081] minF(x) = [f1(x), f2(x),..., f n (x)]

[0082] g j (x) ≤ 0, j = 1, 2,..., m

[0083] h k (x) = 0, k = 1, 2,..., p

[0084] Among them, f i (x) represents the i-th optimization objective.

[0085] In an alternative embodiment, the first adjustment can also be the scheduling optimization of road traffic flow demand. By predicting the future road traffic flow trend, the demand for traffic resources is evaluated, and the traffic resource allocation is adjusted according to the prediction results. Through the feedback of real-time traffic flow data, the actual demand is compared with the predicted demand, and the parameters of the prediction model are adjusted to improve the prediction accuracy. In the process of allocation adjustment, the performance indicators of the accuracy rate, root mean square error, and mean absolute percentage error of the prediction formula are first evaluated, and a real-time scheduling strategy is formulated according to the prediction results. The scheduling decision considers the scheduling strategy at time t and the utility function under the given traffic flow demand. The traffic flow scheduling efficiency index is improved to achieve the optimal allocation and utilization of resources. A multi-objective optimization method is used to improve the performance, and the optimization objectives include prediction accuracy and response time. Constraint conditions are considered to ensure the feasibility and effectiveness of the optimization process, and at the same time, constraint conditions are considered to ensure the feasibility and effectiveness of the optimization process.

[0086] Embodiment 2 is an embodiment of the present invention, which provides an intelligent prediction method for charging demand based on video analysis. In order to verify the beneficial effects of the present invention, scientific demonstrations are carried out through economic benefit calculation and simulation experiments.

[0087] First, a simulation experiment is carried out based on the MATLAB R2023a platform. The actual operation data of a certain city's charging stations for 3 months (including data samples of 1000 vehicles and 10 charging stations) is used as the verification set. In the simulation process, the collected video data is first preprocessed, including adaptive histogram equalization and improved Gaussian filtering. Subsequently, vehicle detection and recognition are realized through an improved YOLOv5 algorithm, and a mapping relationship between vehicle models and charging demands is established. In the spatio-temporal sequence prediction stage, a dual attention mechanism is used for feature extraction and fusion. By comparing with traditional methods such as LSTM, Prophet, and ARIMA, the effectiveness of the algorithm is verified.

[0088] Table 1 Comparison experiment of prediction accuracy

[0089] Evaluation Index The Algorithm in This Paper LSTM Prophet ARIMA Accuracy Rate (%) 94.5 88.2 85.7 82.3 RMSE 0.156 0.243 0.287 0.312 MAPE (%) 6.8 11.5 13.2 15.7

[0090] Table 2 Response time analysis (unit: ms)

[0091] Processing Stage Average Time Maximum Time Minimum Time Image Preprocessing 25.3 35.6 18.2 Vehicle Detection 45.6 58.9 38.4 Demand Forecasting 32.8 42.3 28.1 Total Response Time 103.7 136.8 84.7

[0092] Table 3 Prediction performance under different scenarios

[0093] Scene Type Accuracy Rate (%) RMSE MAPE (%) Peak Hours on Weekdays 93.2 0.168 7.2 Off-Peak Hours on Weekdays 95.8 0.143 6.1 Weekends 94.1 0.162 6.9 Holidays 91.5 0.185 8.4

[0094] Table 4 System optimization effect

[0095]

[0096]

[0097] Experimental results show that the algorithm proposed in this paper is superior to the comparative algorithms in terms of prediction accuracy, RMSE and MAPE, with an accuracy of 94.5% and an average response time of 103.7ms. It also shows strong adaptability in different scenarios. The resource utilization rate after system optimization is increased by 18.3%, which fully proves the feasibility and advancement of the algorithm in practical applications.

[0098] Embodiment 3 is an embodiment of the present invention, which provides a charging demand intelligent prediction system based on video analysis, including an acquisition and processing module, an analysis and prediction module, and an optimization and adjustment module.

[0099] The acquisition and processing module is used to acquire first characteristic data of the first object and pre-process the first characteristic data.

[0100] The analysis and prediction module is used to analyze the first characteristic data to obtain the impact data, combine the historical data with the impact data to predict the future first demand trend and formulate an early warning mechanism.

[0101] The optimization and adjustment module is used to optimize the system scheduling according to the future first demand trend, and to make a first adjustment to the prediction and scheduling link according to the real-time feedback mechanism.

[0102] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0103] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a definable sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or used in conjunction with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0104] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection portion (electronic device) having one or more wirings, a portable computer diskette (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, a computer-readable medium can even be paper or other suitable media on which a program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or other suitable processing as necessary, and then stored in a computer memory.

[0105] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc. It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

[0106] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A charging demand intelligent prediction method based on video analysis, characterized in that: include: Collecting first feature data of a first object, and preprocessing the first feature data; Analyze the first characteristic data to obtain the impact data, combine the historical data with the impact data to predict the future first demand trend and develop an early warning mechanism; The system is scheduled and optimized based on the future first demand trend, and the first adjustment of the forecast scheduling link is made according to the real-time feedback mechanism.

2. The method for intelligent prediction of charging demand based on video analysis according to claim 1, characterized in that: The first characteristic data includes parameter data of the first object during operation.

3. The method for intelligent prediction of charging demand based on video analysis according to claim 2, characterized in that: The impact data includes data that directly affects the overall operation effect when the first object is running; The first requirement includes target values ​​that each parameter data needs to reach when the first object is working.

4. The method for intelligent prediction of charging demand based on video analysis according to claim 3, characterized in that: The scheduling optimization includes allocating and adjusting resource utilization efficiency according to the future first demand trend.

5. The method for intelligent prediction of charging demand based on video analysis according to claim 4, characterized in that: The first adjustment includes obtaining actual demand based on a real-time feedback mechanism and adjusting the accuracy of the prediction model.

6. The method for intelligent prediction of charging demand based on video analysis according to claim 5, characterized in that: Analyzing the first feature data to obtain the influence data includes optimizing the YOLOv5 algorithm, using the optimized YOLOv5 algorithm to analyze the parameter data of the first object when it is running, and detecting and identifying the vehicle type therein; The optimization of the YOLOv5 algorithm includes calculating the contribution of the weight and bias term of each convolution kernel to the model error through a back propagation algorithm, and updating the weight according to this information to achieve specific optimization of the YOLOv5 algorithm; The prediction process includes establishing a prediction model using a dual attention mechanism, capturing the dynamic changes of temporal features and extracting key information of spatial dimensions, and generating the final prediction result through a nonlinear activation function; The early warning mechanism includes dynamically setting the early warning threshold according to the mean, standard deviation and adjustment parameters of historical data to set an adaptive early warning mechanism.

7. The method for intelligent prediction of charging demand based on video analysis according to claim 6, characterized in that: The allocation adjustment process includes calculating the efficiency index of the charging pile and allocating resources according to the efficiency index; The adjustment of the accuracy of the prediction model includes solving the optimal system parameter configuration through a multi-objective optimization algorithm.

8. A system using the charging demand intelligent prediction method based on video analysis as claimed in any one of claims 1 to 7, characterized in that: It includes collection and processing module, analysis and prediction module, and optimization and adjustment module; The acquisition and processing module is used to acquire first feature data of the first object and pre-process the first feature data; The analysis and prediction module is used to analyze the first characteristic data to obtain the impact data, combine the historical data with the impact data to predict the future first demand trend and formulate an early warning mechanism; The optimization and adjustment module is used to optimize the system scheduling according to the future first demand trend, and to make a first adjustment to the prediction and scheduling link according to the real-time feedback mechanism.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images