Electric power big data visual digital platform based on tourist behavior analysis

Abstract

The invention relates to the technical field of digital tourism services, in particular to an electric power big data visual digital platform based on tourist behavior analysis, which realizes precision and dynamics of electric power demand prediction through multi-dimensional analysis and modeling of tourist behaviors, improves the reasonability of electric power resource allocation, and improves the efficiency of tourist resource allocation. The relation between the power consumption mode and tourist activities is recognized through a page ranking algorithm, it is ensured that power demand prediction is more accurate, the flexible scheduling and optimization capacity of power resources is enhanced, modeling is conducted through a support vector machine, dynamic prediction of power demand changes is achieved, and when tourist behaviors fluctuate, the power demand prediction efficiency is improved. The method can quickly respond and adjust the correlation analysis between resource allocation, behavior modes and power demands according to the prediction result, further establishes a network structure through a graph analysis technology, enables the influence evaluation of behavior changes to be more accurate, enhances the prediction reliability, and improves the prediction accuracy through the matching of historical power data and tourist behavior data. And dynamic load optimization is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of digital tourism services, and in particular to a digital platform for visualization of electric power big data based on tourist behavior analysis. Background Art

[0002] The field of digital tourism service technology aims to improve the overall service level and operational efficiency of the tourism industry through the application of modern information technology. The core goal of this field is to achieve intelligent management of tourism resources and personalized optimization of tourist experience. Through data analysis and technical means, digital tourism services can not only improve tourists' satisfaction and participation, but also help the tourism industry accurately predict tourist needs, optimize resource allocation, improve operational efficiency, and reduce energy waste. In addition, it can also improve the competitiveness of tourism companies in the market and promote the sustainable development of the tourism industry through precise marketing methods and intelligent services.

[0003] The purpose of the power big data visualization digital platform based on tourist behavior analysis is to optimize the allocation of power resources, improve power utilization efficiency and service experience by analyzing tourists' behavior data, and combine tourists' power demand and consumption behavior in different places. With the help of data visualization technology, it helps managers understand tourists' demand dynamics in real time and adopt corresponding strategies.

[0004] Existing technologies are unable to effectively handle the complex relationship between tourist behavior and electricity demand. The analysis of tourist behavior remains at a superficial level and is unable to deeply explore the specific impact of tourist activities on electricity fluctuations, resulting in a lack of flexibility in electricity demand forecasting. During peak periods, electricity distribution is often unbalanced, resulting in irrational resource allocation and affecting electricity efficiency. There is a lack of intelligent dynamic adjustment mechanism in power dispatching, and it is impossible to make real-time optimization based on actual demand changes, resulting in a delayed response to system dispatching and an inability to effectively utilize electricity resources during low-peak periods. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a digital platform for visualization of power big data based on visitor behavior analysis.

[0006] In order to achieve the above purpose, the present invention adopts the following technical solution: A digital platform for visualization of power big data based on tourist behavior analysis includes:

[0007] Tourist behavior pattern recognition module: Based on tourist activity information, stay time and activity type, the module identifies the behavior patterns of different types of tourists by stratifying the statistical distribution of tourist stay areas and activity types, calculates the impact of various activities on power demand fluctuations, and generates tourist behavior pattern recognition results;

[0008] Power demand pattern analysis module: Based on the tourist behavior pattern recognition results, by matching historical power data with tourist behavior characteristics, using a page ranking algorithm, identifying the relationship between the power consumption pattern and tourist activities, calculating the fluctuation of power demand caused by behavior changes, and generating power demand pattern analysis results;

[0009] Behavior and demand correlation analysis module: Based on the power demand pattern analysis results, the graph analysis technology is applied to establish a network structure of tourist behavior and power demand changes, and the relationship and influence between nodes are calculated to identify the behavior patterns that affect power load fluctuations, and generate behavior demand correlation analysis results;

[0010] Load forecasting and scheduling module: Based on the behavior demand correlation analysis results, by modeling the relationship between tourist behavior data and power demand changes, using support vector machines to predict the dynamic changes in power demand and future power demand in each region, and generate load forecast results;

[0011] Dynamic load optimization module: based on the load forecast results, combined with regional power demand and load conditions during off-peak hours, optimizes power dispatch, selects an adaptive power load distribution path, and generates a dynamic load optimization plan;

[0012] Energy consumption optimization module: Based on the dynamic load optimization scheme and combined with the feedback information of the energy management system, it adjusts the power supply in real time, controls the power consumption, and generates energy consumption optimization results;

[0013] Power resource allocation strategy module: Based on the energy consumption optimization results, through dynamic monitoring of the real-time power load in each area, a graphical interface is used to display the power demand and allocation situation in each area, the configuration of power resources is adjusted in real time, and the areas with high and low loads are displayed through heat maps, generating a visualization chart of the power resource allocation strategy.

[0014] As a further solution of the present invention, the tourist behavior pattern recognition module includes an activity area statistics submodule, an activity type distribution submodule and a behavior pattern recognition submodule, wherein:

[0015] Activity area statistics submodule: extracts tourist stay area data based on tourist activity information, stay time and activity type, and obtains the total stay time, stay frequency and activity density of tourists in each area by aggregating the stay time of tourists in each area, and generates stay area statistics;

[0016] Activity type distribution submodule: based on the tourist activity type data, count the participation frequency of different tourists in each activity type, and obtain the distribution ratio of each activity type by weighted average of the duration of each type of activity, and generate the activity type distribution result;

[0017] Behavior pattern recognition submodule: Based on the statistical data of the stay area and the distribution results of activity types, a hierarchical analysis is performed on the activity behaviors of tourists. Combined with the stay patterns and activity distribution of tourists in each area, the behavior patterns of tourists are identified through cluster analysis to generate tourist behavior pattern recognition results.

[0018] As a further solution of the present invention, the power demand pattern analysis module includes a power demand matching submodule, a behavior fluctuation analysis submodule and a power demand pattern generation submodule, wherein:

[0019] Power demand matching submodule: Based on the tourist behavior pattern recognition results, the historical power consumption data is matched, the tourist behavior pattern is compared with the power demand data using time series analysis, the characteristics of power demand fluctuations are extracted, and power demand matching data is generated;

[0020] Behavior fluctuation analysis submodule: based on the power demand matching data, calculate the power demand fluctuation caused by the change of tourists' behavior, divide the time period in combination with historical data, analyze the specific impact of the behavior change on the power demand, and generate the behavior fluctuation analysis results;

[0021] Power demand pattern generation submodule: Based on the behavior fluctuation analysis results, a page ranking algorithm is used to analyze the correlation between the fluctuation patterns of tourist activities and the changes in power demand, a power demand pattern is constructed, and data fitting is performed to generate power demand pattern analysis results.

[0022] As a further solution of the present invention, the page ranking algorithm is according to the formula:

[0023]

[0024] Where: P(i) represents the weight of the i-th tourist activity, d represents the damping factor, N represents the total number of activities, B(i) represents the set of other activities pointing to the i-th activity, L(j) represents the number of links to the j-th activity, α represents the influence coefficient of the activity duration, T(i) represents the duration of the i-th activity, β represents the influence coefficient of the activity type, and A(i) represents the type score of the i-th activity.

[0025] As a further solution of the present invention, the behavior and demand correlation analysis module includes a network structure establishment submodule, a relationship calculation submodule and an impact behavior identification submodule, wherein:

[0026] Network structure establishment submodule: Based on the power demand pattern analysis results, the tourist behavior pattern and power demand change data are extracted and processed as nodes for network processing, and the node connection relationship between the tourist behavior pattern and the power demand fluctuation is established. By dividing the node relationship, the power demand network structure is generated;

[0027] Relationship calculation submodule: based on the power demand network structure, calculate the weights between nodes, analyze the correlation strength between tourist behavior patterns and power demand by weighted evaluation of the influence of each node, calculate the influence relationship between nodes, and generate node relationship analysis results;

[0028] Influencing behavior identification submodule: Based on the node relationship analysis results, the tourist behavior patterns significantly related to power load fluctuations are identified, the key influencing factors of the behavior patterns on the changes in power demand are extracted, and the behavior-demand correlation analysis results are generated.

[0029] As a further solution of the present invention, the load prediction and scheduling module includes a modeling analysis submodule, a dynamic prediction submodule and a load scheduling generation submodule, wherein:

[0030] Modeling and analysis submodule: Based on the behavior demand correlation analysis results, fit the relationship between tourist behavior data and electricity demand changes, collect tourist behavior data from various regions, and compare them with electricity demand fluctuation data to generate a demand change relationship model;

[0031] Dynamic prediction submodule: Based on the demand change relationship model, support vector machine is used to dynamically predict the future power demand of each region, and by analyzing the trend of historical power demand data, a change curve of future power demand of each region is generated to generate a dynamic prediction result of power demand;

[0032] Load dispatch generation submodule: Based on the dynamic forecast results of power demand, analyze the future power load of each region, adjust the power dispatch plan of each region according to the power demand forecast data, and generate load forecast results.

[0033] As a further solution of the present invention, the support vector machine is according to the formula:

[0034]

[0035] in: represents the predicted power demand output, w ′ represents the weight vector of the support vector machine, x represents the input feature vector, b ′represents the bias term, α represents the influence coefficient of activity duration on power demand forecast, T(x) represents the duration of the activity, β represents the influence coefficient of weather conditions on power demand forecast, P(x) represents external climate conditions, γ represents the influence coefficient of population density on power demand forecast, and H(x) represents the population density data of a specific area.

[0036] As a further solution of the present invention, the dynamic load optimization module includes a load scheduling optimization submodule, a load path selection submodule and a dynamic optimization solution generation submodule, wherein:

[0037] Load dispatch optimization submodule: Based on the load forecast results, analyze the power demand of each region and the load situation during non-peak hours, and adjust the power dispatch plan for each period in combination with the actual load fluctuation data to reduce the overload during peak hours. By adjusting the power distribution in the region, the load dispatch optimization results are generated;

[0038] Load path selection submodule: Based on the load scheduling optimization results, the power load transmission paths of each region are evaluated, the most suitable power transmission path is selected, and the rationality of power load distribution is ensured by analyzing the capacity and load matching of the path, and the power load distribution path is generated;

[0039] Dynamic optimization scheme generation submodule: Based on the power load distribution path, combined with the real-time power demand of each region, the scheduling scheme is optimized to ensure the continuity and stability of power distribution during load fluctuations, and a dynamic load optimization scheme is generated.

[0040] As a further solution of the present invention, the energy consumption optimization module includes a feedback data collection submodule, a real-time adjustment submodule and an energy consumption control submodule, wherein:

[0041] Feedback data collection submodule: Based on the dynamic load optimization scheme, real-time power supply and consumption data is obtained from the energy management system, and the power consumption data of each area is extracted by collecting system feedback information, and preliminary processing is performed to generate a feedback data set;

[0042] Real-time adjustment submodule: based on the feedback data set, analyze the difference between power supply and demand, adjust the real-time supply of power by adjusting the power supply, optimize the load balance of the power system, and generate real-time adjustment data;

[0043] Energy consumption control submodule: Based on the real-time adjustment data, the power consumption of each area is controlled, the gap between power supply and consumption is adjusted in real time, the power consumption is ensured to be within a predetermined range, and the energy consumption optimization result is generated.

[0044] As a further solution of the present invention, the power resource allocation strategy module includes a load monitoring submodule, a resource adjustment submodule and a visual display submodule, wherein:

[0045] Load monitoring submodule: based on the energy consumption optimization results, collect real-time power load data of each region, analyze the power consumption trends of different regions, monitor changes in power demand, record data and track regional load changes, and generate real-time power load monitoring data;

[0046] Resource adjustment submodule: Based on the real-time power load monitoring data and combined with the power demand of each region, analyze the difference between the loads of each region, optimize the power configuration by adjusting the allocation ratio of power resources, ensure the balance of power loads in each region, and generate a power resource adjustment plan;

[0047] Visual display submodule: Based on the power resource adjustment plan, a graphical interface is used to display the power demand and distribution situation of each region, a heat map is generated, and high and low load areas are distinguished by different colors. The power resource allocation situation is displayed through a visual chart, and a visual chart of the power resource allocation strategy is generated.

[0048] Compared with the prior art, the advantages and positive effects of the present invention are:

[0049] 1. In the present invention, through multi-dimensional analysis and modeling of tourist behavior, the power demand forecast is made accurate and dynamic, and the rationality of power resource allocation is improved;

[0050] 2. In the present invention, the relationship between power consumption patterns and tourist activities is identified through a page ranking algorithm, which ensures more accurate power demand forecasts and enhances the flexible scheduling and optimization capabilities of power resources;

[0051] 3. In the present invention, the modeling is carried out by support vector machine, which realizes the dynamic prediction of the change of power demand. When the behavior of tourists fluctuates, it can quickly respond and adjust the resource allocation according to the prediction results;

[0052] 4. In the present invention, the correlation analysis between behavior patterns and power demand further establishes a network structure through graph analysis technology, making the impact assessment of behavior changes more accurate, enhancing the reliability of predictions, and achieving dynamic load optimization by matching historical power data with tourist behavior data. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 is a system flow chart of the present invention;

[0054] Figure 2 is a flow chart of the present invention;

[0055] Figure 3It is a schematic diagram of the system framework of the present invention. DETAILED DESCRIPTION

[0056] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0057] See also Figure 1 and Figure 2 The present invention provides a technical solution: a digital platform for visualization of power big data based on tourist behavior analysis, comprising:

[0058] Tourist behavior pattern recognition module: Based on tourist activity information, stay time and activity type, the module identifies the behavior patterns of different types of tourists by stratifying the statistical distribution of tourist stay areas and activity types, calculates the impact of various activities on power demand fluctuations, and generates tourist behavior pattern recognition results;

[0059] Power demand pattern analysis module: Based on the identification results of tourist behavior patterns, by matching historical power data with tourist behavior characteristics, using page ranking algorithm, identifying the relationship between power consumption patterns and tourist activities, calculating the fluctuation of power demand caused by behavioral changes, and generating power demand pattern analysis results;

[0060] Behavior and demand correlation analysis module: Based on the analysis results of power demand patterns, graph analysis technology is used to establish a network structure of tourist behavior and changes in power demand. By calculating the relationship and influence between nodes, the behavior patterns that affect power load fluctuations are identified, and the behavior-demand correlation analysis results are generated;

[0061] Load forecasting and scheduling module: Based on the results of behavior demand correlation analysis, the relationship between tourist behavior data and power demand changes is modeled, and support vector machines are used to predict the dynamic changes in power demand and the future power demand of each region, generating load forecast results;

[0062] Dynamic load optimization module: Based on the load forecast results, combined with regional power demand and load conditions during off-peak hours, it optimizes power dispatch, selects an appropriate power load distribution path, and generates a dynamic load optimization plan;

[0063] Energy consumption optimization module: Based on the dynamic load optimization scheme and combined with the feedback information from the energy management system, it adjusts the power supply in real time, controls the power consumption, and generates energy consumption optimization results;

[0064] Power resource allocation strategy module: Based on the energy consumption optimization results, through dynamic monitoring of the real-time power load in each region, a graphical interface is used to display the power demand and allocation situation in each region, the configuration of power resources is adjusted in real time, and the areas with high and low loads are displayed through heat maps, generating a visualization chart of the power resource allocation strategy.

[0065] See also Figure 3 ,The tourist behavior pattern recognition module includes the activity area statistics submodule, the activity type distribution submodule and the behavior pattern recognition submodule, where:

[0066] Activity area statistics submodule: extracts tourist stay area data based on tourist activity information, stay time and activity type, and obtains the total stay time, stay frequency and activity density of tourists in each area by aggregating the stay time of tourists in each area, and generates stay area statistics;

[0067] Activity type distribution submodule: Based on the tourist activity type data, the participation frequency of different tourists in each activity type is counted, and the distribution ratio of each activity type is obtained by weighted average of the duration of each type of activity to generate the activity type distribution result;

[0068] Behavior pattern recognition submodule: Based on the statistical data of the stay area and the distribution results of the activity types, the activity behaviors of tourists are analyzed hierarchically. Combined with the stay patterns and activity distribution of tourists in each area, the behavior patterns of tourists are identified through cluster analysis to generate the recognition results of tourist behavior patterns.

[0069] Activity area statistics submodule: Based on tourist activity information, stay time and activity type, extract tourist stay area data, aggregate and calculate the stay time of tourists in each area, use aggregation function to sum the stay time of tourists in each area, use counting function to count the frequency of tourists staying in each area, and use activity density calculation formula to calculate the activity density in each area. Specifically, the stay area statistics are generated by calculating the ratio of the total stay time of tourists in the area to the area of ​​the area. For each area, the specific values ​​of the total stay time, stay frequency and activity density are generated;

[0070] Activity type distribution submodule: Based on the tourist activity type data, the frequency of participation of different tourists in each activity type is counted, the frequency distribution function is used to count different activity types, and the weighted average algorithm is used to weight the duration of each activity type. When calculating specifically, by setting the activity weight value, the duration of each activity is multiplied by its weight, and then the weighted values ​​of all activities are summed and divided by the total number of activities to generate the activity type distribution result. This step includes the calculation of the frequency distribution and weighted average duration of each activity type;

[0071] Behavior pattern recognition submodule: Based on the statistical data of the stay area and the distribution results of the activity types, a hierarchical analysis of the activity behaviors of tourists is performed, and the K-means clustering algorithm is used to identify the behavior patterns of tourists. During the specific implementation, the initial cluster center is set, and the distance between each tourist and each cluster center is calculated by Euclidean distance. According to the minimum distance principle, tourists are assigned to the nearest cluster center, and iterative steps are performed until the cluster centers converge. Finally, the behavior pattern corresponding to each tourist is obtained, and the tourist behavior pattern recognition results are generated. This step includes clustering processing of the stay pattern and activity distribution of each area, combined with the relationship between the behavior pattern and the change in electricity demand, to ensure accurate identification of tourist behavior patterns.

[0072] See also Figure 3 The power demand pattern analysis module includes a power demand matching submodule, a behavior fluctuation analysis submodule and a power demand pattern generation submodule, wherein:

[0073] Power demand matching submodule: Based on the results of tourist behavior pattern recognition, historical power consumption data is matched, and time series analysis is used to compare tourist behavior patterns with power demand data, extract the characteristics of power demand fluctuations, and generate power demand matching data;

[0074] Behavior fluctuation analysis submodule: Based on the power demand matching data, the power demand fluctuation caused by the change of tourists' behavior is calculated, and the time period is divided according to the historical data to analyze the specific impact of the behavior change on the power demand and generate the behavior fluctuation analysis results;

[0075] Power demand pattern generation submodule: Based on the results of behavior fluctuation analysis, the page ranking algorithm is used to analyze the correlation between the fluctuation pattern of tourist activities and the changes in power demand, construct a power demand pattern, and perform data fitting to generate power demand pattern analysis results;

[0076] Power demand matching submodule: Based on the results of tourist behavior pattern recognition, historical power consumption data is matched, and the tourist behavior pattern is compared with the power demand data using the time series analysis method. First, the timestamps in the tourist behavior pattern are extracted, and the behavior pattern is aligned with the historical power consumption data using the sliding window method. The characteristics of power demand fluctuations are extracted, and power demand matching data is generated. Finally, the relationship between power demand and tourist behavior is predicted through the model;

[0077] Behavior fluctuation analysis submodule: Based on the power demand matching data, the power demand fluctuation caused by the change of tourists' behavior is calculated. The time period division method is adopted. First, the time period is divided according to the fluctuation characteristics of historical data. The weighted average method is used to analyze the specific impact of tourists' behavior on power demand in each time period.

[0078] Set the weighting coefficient, perform weighted processing on each behavior fluctuation, and calculate the impact value of the weighted behavior on electricity demand. When executing, use the weighted average formula to multiply the behavior data in each time period by the weight, and then sum all the weighted values ​​to generate the behavior fluctuation analysis results. The results include the specific fluctuation impact of the behavior changes in each time period on electricity demand;

[0079] Power demand pattern generation submodule: Based on the behavior fluctuation analysis results, the page ranking algorithm is used to first associate the behavior fluctuation data with the power demand data. The contribution of each tourist behavior fluctuation to the power demand is calculated through the page ranking algorithm. The initial weight value is set to 1, and the mutual influence weights between the nodes are calculated. According to the iterative update formula, the weight value of each behavior fluctuation is continuously adjusted until the algorithm converges and generates the power demand pattern analysis results. This process involves fitting the volatility of power demand fluctuations and using the least squares method for fitting analysis to ensure that an accurate power demand pattern is obtained, and finally the power demand pattern analysis results are output.

[0080] Page ranking algorithm, according to the formula:

[0081]

[0082] Where: P(i) represents the weight of the i-th tourist activity, d represents the damping factor, N represents the total number of activities, B(i) represents the set of other activities pointing to the i-th activity, L(j) represents the number of links to the j-th activity, α represents the influence coefficient of the activity duration, T(i) represents the duration of the i-th activity, β represents the influence coefficient of the activity type, and A(i) represents the type score of the i-th activity;

[0083] Implementation process: First, analyze the behavior of tourists. For each activity i, calculate P(i) to measure the impact of the activity on electricity demand. P(i) is calculated through two main parts. The first part represents a basic distribution, which is used to consider the initial weight of each activity, where N is the total number of activities and d is the damping factor used to adjust the probability of random jumps. The second part By weighting the impact of activities related to activity i, the impact of activity on the electricity demand of activity i is considered. P(j) represents the weight of activity j, L(j) is the number of links between activity j and other activities, and by adding new terms α·log(T(i)) and β·A(i), T(i) is the duration of activity i, A(i) is the score of the activity type, and weighting is performed by corresponding coefficients α and β, which reflects the different impacts of different activity durations and types on electricity demand. Finally, by comprehensively considering the connection relationship, duration and type of activities, a more accurate electricity demand pattern is obtained, which provides detailed analysis results for the power big data visualization platform.

[0084] See also Figure 3 The behavior and demand correlation analysis module includes a network structure establishment submodule, a relationship calculation submodule and an impact behavior identification submodule, among which:

[0085] Network structure establishment submodule: Based on the analysis results of power demand patterns, the tourist behavior patterns and power demand change data are extracted and processed as nodes for networking. The node connection relationship between tourist behavior patterns and power demand fluctuations is established. By dividing the node relationship, the power demand network structure is generated.

[0086] Relationship calculation submodule: Based on the power demand network structure, the weights between nodes are calculated. By weighted evaluation of the influence of each node, the correlation strength between tourist behavior patterns and power demand is analyzed, the influence relationship between nodes is calculated, and the node relationship analysis results are generated;

[0087] Influencing behavior identification submodule: Based on the node relationship analysis results, it identifies the tourist behavior patterns that are significantly related to power load fluctuations, extracts the key influencing factors of the behavior patterns on power demand changes, and generates behavior-demand correlation analysis results;

[0088] Network structure establishment submodule: Based on the results of power demand pattern analysis, tourist behavior patterns and power demand change data are extracted and networked as nodes. A graph construction algorithm is used to establish a node connection relationship between each tourist behavior pattern and power demand fluctuation through an adjacency matrix. The Dijkstra algorithm is used to calculate the shortest path between nodes. The node relationship is divided according to the correlation and distance between nodes to generate a power demand network structure. This process includes setting the weight value of the node connection, determining the connection strength between nodes by setting the weight coefficient, and generating a structured power demand network diagram.

[0089] Relationship calculation submodule: Based on the power demand network structure, the weights between nodes are calculated. The specific steps include initializing the weight of each node to 1, applying the iterative update formula to each node in the network, adjusting the weight value through the connection relationship between nodes, setting the damping coefficient to 0.85, iterating until the node weights converge, calculating the influence of each node and performing a weighted evaluation, using the weighted calculation formula to weight the influence of each node with its weight, analyzing the correlation strength between tourist behavior patterns and power demand, and finally calculating the influence relationship between nodes to generate node relationship analysis results;

[0090] Influencing behavior identification submodule: Based on the results of node relationship analysis, the tourist behavior patterns that are significantly related to power load fluctuations are identified. The cluster analysis method is adopted, and the K-means algorithm is used to cluster the node relationship analysis results. The number of clusters is set to k = 3, and the Euclidean distance is selected as the distance metric. The iterative method is used for convergence processing to identify the tourist behavior patterns that are significantly related to power load fluctuations, extract the key influencing factors of the behavior patterns on the changes in power demand, and finally generate the behavior-demand correlation analysis results.

[0091] See also Figure 3 The load prediction and scheduling module includes a modeling and analysis submodule, a dynamic prediction submodule, and a load scheduling generation submodule, among which:

[0092] Modeling and analysis submodule: Based on the analysis results of the correlation between behavior and demand, the relationship between tourist behavior data and electricity demand changes is fitted, tourist behavior data from various regions is collected, and compared with electricity demand fluctuation data to generate a demand change relationship model;

[0093] Dynamic prediction submodule: Based on the demand change relationship model, support vector machine is used to dynamically predict the future power demand of each region. By analyzing the trend of historical power demand data, the change curve of future power demand in each region is generated, and the dynamic prediction result of power demand is generated;

[0094] Load dispatch generation submodule: Based on the dynamic forecast results of power demand, analyze the future power load of each region, adjust the power dispatch plan of each region according to the power demand forecast data, and generate load forecast results;

[0095] Modeling and analysis submodule: Based on the analysis results of the correlation between behavior and demand, the relationship between tourist behavior data and electricity demand changes is fitted. Tourist behavior data from various regions are collected and compared with electricity demand fluctuation data. The least squares method is used for fitting analysis. Specifically, the loss function is set as the square error, and the residual sum of squares between tourist behavior data and electricity demand fluctuation data is calculated. By minimizing the loss function, the parameters of the behavior demand relationship model are obtained, and the demand change relationship model is generated. This process includes adjusting the parameters of the fitting function to make the prediction results more consistent with the actual demand change trend;

[0096] Dynamic prediction submodule: Based on the demand change relationship model, support vector machine is used for dynamic prediction. First, a training data set is prepared, the C value is set to 1.0, the radial basis kernel function is selected as the kernel function of the support vector machine, and the γ value is set to 0.5. The support vector machine model is trained by trend analysis of historical power demand data, and the trained model is used to predict the future power demand of each region. The prediction results are tuned by support vector machine regression, and the change curve of future power demand in each region is generated. The dynamic prediction results of power demand are generated. This process includes calculating the prediction error of the support vector machine model and adjusting it to improve the accuracy of the prediction;

[0097] Load scheduling generation submodule: Based on the dynamic prediction results of power demand, the future power load of each region is analyzed, and the optimization algorithm is used to adjust the power scheduling plan of each region. Specifically, the particle swarm optimization algorithm is used, the number of particles is set to 30, and the maximum number of iterations is set to 1000. In each iteration, the fitness value of the particle position is calculated to optimize the objective function of power load scheduling, update the particle position, and find the optimal power scheduling plan through a combination of local search and global search to generate load prediction results. This process includes updating the power scheduling strategy through the particle swarm algorithm and optimizing the future power supply plan of each region.

[0098] Support vector machine, according to the formula:

[0099]

[0100] in: represents the predicted power demand output, w ′ represents the weight vector of the support vector machine, x represents the input feature vector, b ′ represents the bias term, α represents the influence coefficient of activity duration on power demand forecast, T(x) represents the duration of the activity, β represents the influence coefficient of weather conditions on power demand forecast, P(x) represents the external climate conditions, γ represents the influence coefficient of population density on power demand forecast, and H(x) represents the population density data of a specific area;

[0101] Execution process: First, by collecting characteristic data related to tourist behavior, construct the input feature vector x, and then perform model calculation. The model is calculated through the weight vector w ′ The inner product of the input feature vector x, combined with the bias term b ′, and obtain the preliminary power demand forecast value. Then, multiple additional factors are added. The influence of activity duration is represented by the term α·T(x), where T(x) is the duration of the activity. The coefficient α is obtained by regression analysis of historical data, reflecting the impact of activity time on power demand. Then, the influence of external climate conditions is added. The term β·P(x) is used to represent the correction of environmental factors such as temperature and humidity on power demand. The coefficient β is also determined by regression analysis of historical data. Finally, in order to consider the impact of regional population, the term γ·H(x) is added. H(x) represents the population density of the region. The coefficient γ is obtained by regression analysis, indicating the impact of population distribution on power demand. Finally, a more accurate power demand forecast value is obtained, providing an accurate power demand dynamic change curve for the power big data visualization platform.

[0102] See also Figure 3 The dynamic load optimization module includes a load scheduling optimization submodule, a load path selection submodule and a dynamic optimization solution generation submodule, wherein:

[0103] Load dispatch optimization submodule: Based on the load forecast results, analyze the power demand of each region and the load situation during non-peak hours. Combined with the actual load fluctuation data, adjust the power dispatch plan for each period to reduce the overload during peak hours. By adjusting the power distribution in the region, the load dispatch optimization results are generated.

[0104] Load path selection submodule: Based on the load scheduling optimization results, the power load transmission paths in each region are evaluated, and the most suitable power transmission path is selected. By analyzing the capacity and load matching of the path, the rationality of power load distribution is ensured, and the power load distribution path is generated;

[0105] Dynamic optimization scheme generation submodule: Based on the power load distribution path and combined with the real-time power demand of each region, the scheduling scheme is optimized to ensure the continuity and stability of power distribution during load fluctuations and generate a dynamic load optimization scheme;

[0106] Load dispatch optimization submodule: Based on the load forecast results, the power demand of each region and the load situation during non-peak hours are analyzed. Combined with the actual load fluctuation data, the linear programming algorithm is used to optimize the power dispatch scheme. First, the objective function is set to minimize the possibility of load overload during peak hours. The constraints include the power demand of each time period and the load fluctuation range during non-peak hours. The Simplex method is used to solve the linear programming problem, calculate the power dispatch demand of each region in different time periods, and generate load dispatch optimization results by adjusting the power distribution in the region. This process includes the adjustment of power distribution in each time period to reduce the risk of load overload during peak hours and ensure the rationality of power distribution in each time period.

[0107] Load path selection submodule: Based on the load scheduling optimization results, the power load transmission paths of each region are evaluated, and the shortest path algorithm, specifically the Dijkstra algorithm, is used to select the most suitable power transmission path in the power load transmission network, and the capacity parameters and load matching degree of each path are set. The transmission capacity of the calculated path is matched with the load demand of each region to ensure the rationality of power load distribution and generate the power load distribution path. This process includes the evaluation of each path and the selection of the optimal transmission path with matching capacity and load;

[0108] Dynamic optimization scheme generation submodule: Based on the power load distribution path and combined with the real-time power demand of each region, a genetic algorithm is used to optimize the scheduling scheme. The population size is set to 50, the crossover rate is selected to be 0.8, and the mutation rate is 0.1. The population is initialized as a set of randomly generated scheduling schemes, and crossover and mutation operations are performed in each generation. By calculating the fitness function, the effectiveness of each scheme is evaluated, the scheduling scheme is gradually optimized, and a dynamic load optimization scheme is generated. This process includes optimizing the continuity and stability of power distribution during load fluctuations to ensure that the power scheduling scheme can adapt to real-time changing needs.

[0109] See also Figure 3 ,The energy consumption optimization module includes a feedback data collection submodule, a real-time adjustment submodule and an energy consumption control submodule, where:

[0110] Feedback data collection submodule: Based on the dynamic load optimization solution, it obtains real-time power supply and consumption data from the energy management system, extracts power consumption data of each area by collecting system feedback information, and performs preliminary processing to generate feedback data sets;

[0111] Real-time adjustment submodule: Based on the feedback data set, it analyzes the difference between power supply and demand, adjusts the real-time supply of power by adjusting the power supply, optimizes the load balance of the power system, and generates real-time adjustment data;

[0112] Energy consumption control submodule: Based on real-time adjustment data, it controls the power consumption of each area, adjusts the gap between power supply and consumption in real time, ensures that power consumption is within the predetermined range, and generates energy consumption optimization results;

[0113] Feedback data collection submodule: Based on the dynamic load optimization solution, real-time power supply and consumption data is obtained from the energy management system. By collecting system feedback information, the power consumption data of each area is obtained using the API interface. Specifically, the feedback information is parsed in JSON format to extract the power consumption and supply data of each area. The raw data is pre-processed using a data denoising algorithm, and the data is smoothed using a moving average filter to remove sudden interference signals and generate a feedback data set. This process includes timestamp synchronization of the raw power data to ensure that the power data of each area corresponds to its actual demand time period;

[0114] Real-time adjustment submodule: Based on the feedback data set, analyze the difference between power supply and demand, calculate the difference between power supply and demand through the difference method, and use the PID control algorithm to adjust the power supply in real time. First, set the proportional coefficient Kp to 1.5, the integral coefficient Ki to 0.1, and the differential coefficient Kd to 0.01. Calculate the error based on the difference between the current power supply and demand, output the adjustment amount, adjust the real-time supply of the power system, control the power load in real time, and generate real-time adjustment data. This process includes comparing the power demand with the actual supply, implementing the adjustment plan to reduce the load deviation, and ensuring the stability of the power supply;

[0115] Energy consumption control submodule: Based on real-time adjustment data, the power consumption of each area is controlled. The linear programming method is used to set the objective function to minimize the gap between power supply and consumption. Specifically, the upper and lower limits of power consumption in each area are defined through constraints. The simplex method is selected to solve the optimization problem, calculate and adjust the power supply of each area, adjust the gap between power supply and consumption in real time, and generate energy consumption optimization results. This process includes adjusting power distribution through optimization algorithms to ensure that power consumption is within a predetermined range and reduce unnecessary energy waste.

[0116] See also Figure 3 ,The power resource allocation strategy module includes a load monitoring submodule, a resource adjustment submodule and a visual display submodule, where:

[0117] Load monitoring submodule: Based on the energy consumption optimization results, collect real-time power load data of each region, analyze the power consumption trends of different regions, monitor changes in power demand, record data and track regional load changes, and generate real-time power load monitoring data;

[0118] Resource adjustment submodule: Based on real-time power load monitoring data and combined with the power demand of each region, analyze the differences between loads in each region, optimize power configuration by adjusting the allocation ratio of power resources, ensure the balance of power load in each region, and generate a power resource adjustment plan;

[0119] Visualization display submodule: Based on the power resource adjustment plan, a graphical interface is used to display the power demand and distribution of each region, a heat map is generated, and high and low load areas are distinguished by different colors. The power resource allocation is displayed through a visual chart, and a visual chart of the power resource allocation strategy is generated;

[0120] Load monitoring submodule: Based on the results of energy consumption optimization, collect real-time power load data of each region, use time series analysis to model the power load change trend, use ARIMA model to fit the power load data, specifically set the p value to 2, d value to 1, q value to 2, predict the trend of power load changes in each region, analyze the power consumption trend in different regions, monitor the change of power demand, use data recording function to record the power load data of each period, track regional load changes, and generate real-time power load monitoring data. This process includes monitoring the real-time data of each region moment by moment, generating load change curves and making dynamic adjustments;

[0121] Resource adjustment submodule: Based on real-time power load monitoring data and combined with the power demand of each region, the difference analysis method is used to calculate the difference between the loads of each region. Specifically, the power loads of different regions are compared using analysis of variance (ANOVA), and the significance level is set to 0.05. By analyzing the load differences between regions, the adjustment priority of power resources is determined. By adjusting the allocation ratio of power resources, the optimization algorithm is used to calculate the best solution for resource reallocation to ensure the balance of power loads in each region and generate a power resource adjustment plan. This process includes weighted evaluation of the power demand of each region and adjustment based on the evaluation results.

[0122] Visual display submodule: Based on the power resource adjustment plan, a heat map generation algorithm is adopted, and a graphical interface is drawn using the matplotlib library. First, the color mapping parameter is set to coolwarm, and color mapping is performed according to the power load of each area. By setting the color scale range to 0-100%, high and low load areas are distinguished by different colors to generate a heat map. Combined with the visualization chart tool, the power resource allocation situation is displayed through the ggplot2 function, and a visualization chart of the power resource allocation strategy is generated. This process includes the graphical display of the specific details of the power resource allocation, making the power demand and allocation situation of each area more intuitive.

[0123] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A digital platform for visualization of power big data based on tourist behavior analysis, characterized in that: The platform includes: Tourist behavior pattern recognition module: Based on tourist activity information, stay time and activity type, the module identifies the behavior patterns of different types of tourists by stratifying the statistical distribution of tourist stay areas and activity types, calculates the impact of various activities on power demand fluctuations, and generates tourist behavior pattern recognition results; Power demand pattern analysis module: Based on the tourist behavior pattern recognition results, by matching historical power data with tourist behavior characteristics, using a page ranking algorithm, identifying the relationship between the power consumption pattern and tourist activities, calculating the fluctuation of power demand caused by behavior changes, and generating power demand pattern analysis results; Behavior and demand correlation analysis module: Based on the power demand pattern analysis results, the graph analysis technology is applied to establish a network structure of tourist behavior and power demand changes, and the relationship and influence between nodes are calculated to identify the behavior patterns that affect power load fluctuations, and generate behavior demand correlation analysis results; Load forecasting and scheduling module: Based on the behavior demand correlation analysis results, by modeling the relationship between tourist behavior data and power demand changes, using support vector machines to predict the dynamic changes in power demand and future power demand in each region, and generate load forecast results; Dynamic load optimization module: based on the load forecast results, combined with regional power demand and load conditions during off-peak hours, optimizes power dispatch, selects an adaptive power load distribution path, and generates a dynamic load optimization plan; Energy consumption optimization module: Based on the dynamic load optimization scheme and combined with the feedback information of the energy management system, it adjusts the power supply in real time, controls the power consumption, and generates energy consumption optimization results; Power resource allocation strategy module: Based on the energy consumption optimization results, through dynamic monitoring of the real-time power load in each area, a graphical interface is used to display the power demand and allocation situation in each area, the configuration of power resources is adjusted in real time, and the areas with high and low loads are displayed through heat maps, generating a visualization chart of the power resource allocation strategy.

2. The electric power big data visualization digital platform based on tourist behavior analysis according to claim 1 is characterized in that: The tourist behavior pattern recognition module includes an activity area statistics submodule, an activity type distribution submodule and a behavior pattern recognition submodule, wherein: Activity area statistics submodule: extracts tourist stay area data based on tourist activity information, stay time and activity type, and obtains the total stay time, stay frequency and activity density of tourists in each area by aggregating the stay time of tourists in each area, and generates stay area statistics; Activity type distribution submodule: based on the tourist activity type data, count the participation frequency of different tourists in each activity type, and obtain the distribution ratio of each activity type by weighted average of the duration of each type of activity, and generate the activity type distribution result; Behavior pattern recognition submodule: Based on the statistical data of the stay area and the distribution results of activity types, a hierarchical analysis is performed on the activity behaviors of tourists. Combined with the stay patterns and activity distribution of tourists in each area, the behavior patterns of tourists are identified through cluster analysis to generate tourist behavior pattern recognition results.

3. The electric power big data visualization digital platform based on tourist behavior analysis according to claim 1 is characterized in that: The power demand pattern analysis module includes a power demand matching submodule, a behavior fluctuation analysis submodule and a power demand pattern generation submodule, wherein: Power demand matching submodule: Based on the tourist behavior pattern recognition results, the historical power consumption data is matched, the tourist behavior pattern is compared with the power demand data using time series analysis, the characteristics of power demand fluctuations are extracted, and power demand matching data is generated; Behavior fluctuation analysis submodule: based on the power demand matching data, calculate the power demand fluctuation caused by the change of tourists' behavior, divide the time period in combination with historical data, analyze the specific impact of the behavior change on the power demand, and generate the behavior fluctuation analysis results; Power demand pattern generation submodule: Based on the behavior fluctuation analysis results, a page ranking algorithm is used to analyze the correlation between the fluctuation patterns of tourist activities and the changes in power demand, a power demand pattern is constructed, and data fitting is performed to generate power demand pattern analysis results.

4. The electric power big data visualization digital platform based on tourist behavior analysis according to claim 3 is characterized in that: The page ranking algorithm is based on the formula: Where: P(i) represents the weight of the i-th tourist activity, d represents the damping factor, N represents the total number of activities, B(i) represents the set of other activities pointing to the i-th activity, L(j) represents the number of links to the j-th activity, α represents the influence coefficient of the activity duration, T(i) represents the duration of the i-th activity, β represents the influence coefficient of the activity type, and A(i) represents the type score of the i-th activity.

5. The electric power big data visualization digital platform based on tourist behavior analysis according to claim 1 is characterized in that: The behavior and demand correlation analysis module includes a network structure establishment submodule, a relationship calculation submodule and an impact behavior identification submodule, wherein: Network structure establishment submodule: Based on the power demand pattern analysis results, the tourist behavior pattern and power demand change data are extracted and processed as nodes for network processing, and the node connection relationship between the tourist behavior pattern and the power demand fluctuation is established. By dividing the node relationship, the power demand network structure is generated; Relationship calculation submodule: based on the power demand network structure, calculate the weights between nodes, analyze the correlation strength between tourist behavior patterns and power demand by weighted evaluation of the influence of each node, calculate the influence relationship between nodes, and generate node relationship analysis results; Influencing behavior identification submodule: Based on the node relationship analysis results, the tourist behavior patterns significantly related to power load fluctuations are identified, the key influencing factors of the behavior patterns on the changes in power demand are extracted, and the behavior-demand correlation analysis results are generated.

6. The electric power big data visualization digital platform based on tourist behavior analysis according to claim 1 is characterized in that: The load prediction and scheduling module includes a modeling analysis submodule, a dynamic prediction submodule and a load scheduling generation submodule, wherein: Modeling and analysis submodule: Based on the behavior demand correlation analysis results, fit the relationship between tourist behavior data and electricity demand changes, collect tourist behavior data from various regions, and compare them with electricity demand fluctuation data to generate a demand change relationship model; Dynamic prediction submodule: Based on the demand change relationship model, support vector machine is used to dynamically predict the future power demand of each region, and by analyzing the trend of historical power demand data, a change curve of future power demand of each region is generated to generate a dynamic prediction result of power demand; Load dispatch generation submodule: Based on the dynamic forecast results of power demand, analyze the future power load of each region, adjust the power dispatch plan of each region according to the power demand forecast data, and generate load forecast results.

7. The electric power big data visualization digital platform based on tourist behavior analysis according to claim 6 is characterized in that: The support vector machine is based on the formula: in: represents the predicted power demand output, w ′ represents the weight vector of the support vector machine, x represents the input feature vector, b ′ represents the bias term, α represents the influence coefficient of activity duration on power demand forecast, T(x) represents the duration of the activity, β represents the influence coefficient of weather conditions on power demand forecast, P(x) represents external climate conditions, γ represents the influence coefficient of population density on power demand forecast, and H(x) represents the population density data of a specific area.

8. The electric power big data visualization digital platform based on tourist behavior analysis according to claim 1 is characterized in that: The dynamic load optimization module includes a load scheduling optimization submodule, a load path selection submodule and a dynamic optimization solution generation submodule, wherein: Load dispatch optimization submodule: Based on the load forecast results, analyze the power demand of each region and the load situation during non-peak hours, and adjust the power dispatch plan for each period in combination with the actual load fluctuation data to reduce the overload during peak hours. By adjusting the power distribution in the region, the load dispatch optimization results are generated; Load path selection submodule: Based on the load scheduling optimization results, the power load transmission paths of each region are evaluated, the most suitable power transmission path is selected, and the rationality of power load distribution is ensured by analyzing the capacity and load matching of the path, and the power load distribution path is generated; Dynamic optimization scheme generation submodule: Based on the power load distribution path, combined with the real-time power demand of each region, the scheduling scheme is optimized to ensure the continuity and stability of power distribution during load fluctuations, and a dynamic load optimization scheme is generated.

9. The electric power big data visualization digital platform based on tourist behavior analysis according to claim 1 is characterized in that: The energy consumption optimization module includes a feedback data collection submodule, a real-time adjustment submodule and an energy consumption control submodule, wherein: Feedback data collection submodule: Based on the dynamic load optimization scheme, real-time power supply and consumption data is obtained from the energy management system, and the power consumption data of each area is extracted by collecting system feedback information, and preliminary processing is performed to generate a feedback data set; Real-time adjustment submodule: based on the feedback data set, analyze the difference between power supply and demand, adjust the real-time supply of power by adjusting the power supply, optimize the load balance of the power system, and generate real-time adjustment data; Energy consumption control submodule: Based on the real-time adjustment data, the power consumption of each area is controlled, the gap between power supply and consumption is adjusted in real time, the power consumption is ensured to be within a predetermined range, and the energy consumption optimization result is generated.

10. The electric power big data visualization digital platform based on tourist behavior analysis according to claim 1 is characterized in that: The power resource allocation strategy module includes a load monitoring submodule, a resource adjustment submodule and a visual display submodule, wherein: Load monitoring submodule: based on the energy consumption optimization results, collect real-time power load data of each region, analyze the power consumption trends of different regions, monitor changes in power demand, record data and track regional load changes, and generate real-time power load monitoring data; Resource adjustment submodule: Based on the real-time power load monitoring data and combined with the power demand of each region, analyze the difference between the loads of each region, optimize the power configuration by adjusting the allocation ratio of power resources, ensure the balance of power loads in each region, and generate a power resource adjustment plan; Visual display submodule: Based on the power resource adjustment plan, a graphical interface is used to display the power demand and distribution situation of each region, a heat map is generated, and high and low load areas are distinguished by different colors. The power resource allocation situation is displayed through a visual chart, and a visual chart of the power resource allocation strategy is generated.

Images