A power supply scheduling method and system for a self-service car washing machine

Abstract

The application relates to a power supply scheduling method and system of a self-service car washing machine, which comprises the following steps: obtaining power consumption data information and order data information of the self-service car washing machine; inputting the power consumption data information and the order data information into a prediction model, so that the prediction model outputs demand prediction information; obtaining power supply load information and electricity price floating information, and generating a power supply scheduling strategy; and adjusting power supply parameters and order reservation strategies of the self-service car washing machine based on the power supply scheduling strategy. According to the application, the power consumption data and order data of the self-service car washing machine in a target network point are obtained and input into a prediction model, so that the accurate prediction of power consumption demand and order quantity in a future period is realized; the dynamic optimization of the power supply scheduling strategy is constructed by combining the power supply load information and real-time electricity price floating data; finally, the equipment power supply parameters and the order reservation scheme are synchronously adjusted based on the strategy, a closed-loop control system is formed, the operation cost is reduced, and the effect of maintaining good user experience is achieved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of power supply scheduling, in particular to a power supply scheduling method and system of a self-service car washing machine. BACKGROUND

[0002] As a new intelligent service facility, self-service car washing machines have been rapidly popularized in cities in recent years, and their power supply scheduling problems have gradually become prominent.

[0003] The power supply control methods commonly used in the industry at present mainly fall into two categories: one is preset power control based on a fixed time table, that is, a fixed operating power is set at a specific time period according to historical experience; the other is simple demand response control, which uniformly reduces the power output of all devices when the power grid load is too high. These traditional methods are simple to implement, but have obvious limitations.

[0004] Firstly, the fixed time table mode cannot adapt to the dynamically changing external environment, for example, when sudden weather changes cause a surge in car washing demand, the preset power supply scheme often results in either waste of power resources or insufficient service capacity; secondly, the existing systems generally lack fine response to electricity price signals, and it is difficult to achieve optimal economic operation under the time-of-use electricity price policy. More importantly, the current technical solutions separate power supply control from order management, resulting in a serious mismatch between power supply capacity and user reservation demand during the peak period of power grid load, which not only affects user experience but also causes energy waste.

[0005] The above problems have become a technical bottleneck restricting the large-scale development of self-service car washing services, and an intelligent power supply scheduling scheme that integrates electricity consumption prediction, electricity price response and demand management is urgently needed. SUMMARY

[0006] In order to solve the above-mentioned defects, the application provides a power supply scheduling method and system of a self-service car washing machine.

[0007] The above-mentioned application objectives of the application are achieved by the following technical solutions:

[0008] A power supply scheduling method of a self-service car washing machine, comprising the steps of:

[0009] obtaining electricity consumption data information and order data information of self-service car washing machines in a target network point;

[0010] inputting the electricity consumption data information and the order data information into a pre-trained prediction model, so that the prediction model outputs demand prediction information;

[0011] obtaining power supply load information and electricity price fluctuation information, and generating a power supply scheduling strategy based on the power supply load information, the electricity price fluctuation information and the demand prediction information;

[0012] The power supply parameters and order reservation strategies for self-service car wash machines are adjusted based on power supply scheduling strategies.

[0013] By adopting the above technical solution, this application obtains the electricity consumption and order data of self-service car wash machines within the target network and inputs them into a pre-trained prediction model to achieve accurate prediction of electricity demand and order volume for future periods. By combining power load information and real-time electricity price fluctuation data, a dynamically optimized power supply scheduling strategy is constructed. Finally, based on this strategy, the equipment power supply parameters and order reservation scheme are adjusted synchronously to form a closed-loop control system. By integrating environmental factors and time series analysis into the prediction model, the shortcomings of fixed timetable models in responding to environmental changes are overcome, enabling the power supply scheme to be dynamically adjusted according to external factors such as weather and holidays, thus avoiding waste of electricity resources while ensuring service capacity. Secondly, by introducing electricity price fluctuation information as a core optimization parameter, a refined response to time-of-use pricing policies is achieved. By intelligently reducing power during high-price periods and improving service capacity during low-price periods, operating costs are significantly reduced. Simultaneously, power supply control and order management are optimized collaboratively. When a peak grid load is predicted, the system adjusts the reservation strategy in advance, guiding users to use the service during off-peak hours through price leverage, which ensures the safe and stable operation of the power grid and maintains a good user experience.

[0014] In a preferred embodiment, this application can be further configured such that: the prediction model includes an environmental data acquisition layer, a feature extraction layer, a time series analysis layer, and a prediction output layer; the step of inputting electricity consumption data and order data information into the pre-trained prediction model, so that the prediction model outputs demand prediction information, includes the following steps:

[0015] The environmental data acquisition layer acquires the current external environmental data and current time data of the target site;

[0016] The feature extraction layer extracts environmental features, time features, electricity consumption features, and order features based on external environment data, current time data, electricity consumption data, and order data.

[0017] The time series analysis layer performs time series analysis based on environmental features, time features, electricity consumption features, and order features, thereby generating time series features including electricity demand and order demand.

[0018] The prediction output layer predicts electricity demand based on time series features and outputs demand prediction information.

[0019] By adopting the above technical solution, this application constructs a multi-level prediction model comprising an environmental data acquisition layer, a feature extraction layer, a time series analysis layer, and a prediction output layer to achieve accurate prediction of the electricity demand and order volume of self-service car wash machines. The environmental data acquisition layer collects real-time external environmental data such as weather conditions and temperature at the target site, as well as specific time information, providing basic data support for prediction. The feature extraction layer extracts predictive feature vectors from environmental data, time information, historical electricity consumption records, and order data, including environmental features reflecting weather impacts, time features reflecting periodic changes, electricity consumption features characterizing equipment operating status, and order features displaying user behavior patterns. The time series analysis layer analyzes environmental features, time features, and electricity consumption features... The system performs time-series analysis on characteristics such as electricity demand and order volume to generate time-series features. The prediction output layer forecasts electricity demand based on these time-series features and outputs forecast information. An environmental data acquisition layer enables the perception of external influencing factors, allowing the forecast results to respond promptly to special circumstances such as sudden weather changes. A feature extraction layer extracts deep-seated correlations from the raw data. A time-series analysis layer performs time-series analysis on multi-dimensional features to capture the changing patterns of electricity load and order volume. Finally, the prediction output layer transforms the time-series analysis results into specific demand forecast values. This collaborative design across all modules improves the professionalism and accuracy of data processing at each stage, providing reliable demand forecasting data for power supply dispatching.

[0020] In a preferred embodiment, this application can be further configured as follows: the time series analysis layer performs time series analysis based on environmental characteristics, time characteristics, electricity consumption characteristics, and order characteristics to generate time series characteristics including electricity demand and order demand, including the following steps:

[0021] The time series analysis layer performs weighted fusion of environmental features, time features, electricity consumption features, and order features to generate a joint feature vector;

[0022] The time series analysis layer performs a periodic-based trend decomposition on the joint feature vector to obtain the long-term trend component and the seasonal periodic component.

[0023] The time series analysis layer performs dynamic modeling based on neural networks on the long-term trend component and the seasonal cycle component to obtain the baseline forecast of electricity load and the order fluctuation pattern.

[0024] The time series analysis layer generates time series features based on baseline electricity load forecasts and order fluctuation patterns.

[0025] By adopting the above technical solution, this application achieves accurate identification of the operational demand of self-service car wash machines by weighted fusion and periodic decomposition of various features through time series analysis. Specifically, environmental features, time features, electricity consumption features, and order features are integrated into a unified joint feature vector through feature weighted fusion, ensuring that various influencing factors are reasonably balanced. Periodic trend decomposition technology is used to decompose the joint feature vector into a trend component reflecting long-term changes and a seasonal component reflecting periodic patterns, so that demand features at different time scales can be processed separately. For the decomposed components, neural networks are used for dynamic modeling, where the long-term trend component is used to predict the basic electricity load, and the seasonal periodic component is used to capture the fluctuation pattern of order volume. Finally, the prediction results of the two components are recombined to form a complete time series feature. The above analysis method effectively improves the prediction accuracy through feature fusion and independent component modeling. In particular, for car wash demand scenarios that are simultaneously affected by long-term trends and seasonal cycles, it can more accurately reflect the electricity consumption and order change patterns of different time periods such as weekdays and holidays, providing a reliable decision-making basis for subsequent power supply scheduling.

[0026] In a preferred embodiment, this application can be further configured as follows: the step of obtaining power supply load information and electricity price fluctuation information, and generating a power supply scheduling strategy based on the power supply load information, electricity price fluctuation information, and demand forecast information, includes the following steps:

[0027] Construct a time-axis mapping table of electricity price fluctuation information, and divide the electricity price sensitive range based on demand forecast information;

[0028] The optimal power supply parameters are solved based on a pre-set optimization algorithm combined with electricity price sensitive range and power supply load information.

[0029] Power supply scheduling strategies are generated based on optimal power supply parameters.

[0030] By adopting the above technical solution, this application achieves intelligent generation of power supply dispatching strategies by constructing a time-axis mapping table of electricity price fluctuation information and dividing electricity price sensitive intervals in conjunction with demand forecasts: Time-of-use electricity price data is structured and mapped according to the time axis to establish a precise correspondence between electricity price and time points; based on demand forecast information, different time periods are divided into high-sensitivity intervals and low-sensitivity intervals, where high-sensitivity intervals correspond to periods with higher electricity prices and larger predicted demand, and low-sensitivity intervals are the opposite; on this basis, a preset optimization algorithm is used to comprehensively consider the results of the electricity price sensitive interval division and real-time power supply load information, and the optimal power supply parameters are obtained through mathematical modeling and constraint solving, including the optimal power setting and equipment operation scheme for each time period; an executable power supply dispatching strategy is generated based on the optimal power supply parameters obtained; the above method, through intelligent division of electricity price sensitive intervals and precise solution of optimization algorithms, achieves economical optimized dispatching under the condition of satisfying grid load constraints, avoiding the subjectivity and lag of manually formulated dispatching strategies, and dynamically adjusting the power supply scheme according to electricity price fluctuations and demand changes, thus improving the accuracy and economy of power supply dispatching for self-service car wash machines.

[0031] In a preferred embodiment, this application can be further configured as follows: the electricity price sensitive range includes a high-sensitivity range and a low-sensitivity range; the optimal power supply parameters include high-sensitivity power supply parameters and low-sensitivity power supply parameters; the step of solving for the optimal power supply parameters based on a pre-set optimization algorithm combined with the electricity price sensitive range and power supply load information includes the following steps:

[0032] The high-sensitivity power supply parameters are obtained by combining a pre-set mixed integer programming algorithm with information on high-sensitivity intervals and power supply load.

[0033] The low-sensitivity power supply parameters are obtained by combining a pre-set improved greedy algorithm with low-sensitivity intervals and power supply load information.

[0034] By adopting the above technical solution, this application achieves differentiated scheduling for different electricity price sensitive periods through a dual-mode optimization strategy using a mixed integer programming algorithm and an improved greedy algorithm: For the highly sensitive period, a mixed integer programming algorithm is used for precise modeling and solving, treating equipment start-up and shutdown states as discrete variables and power parameters as continuous variables. Under the premise of satisfying grid load constraints, the highly sensitive power supply parameters that balance economy and reliability are obtained. For the low-sensitivity period, an improved greedy algorithm is used for rapid optimization. This algorithm obtains an approximate optimal solution with low computational cost while ensuring basic power supply needs through the principles of electricity price ranking and capacity priority allocation. The above optimization method not only ensures the accuracy of scheduling during highly sensitive periods but also improves the computational efficiency during low-sensitivity periods, enabling the rapid generation of optimal power supply schemes in complex and ever-changing operating environments. Through precise matching of the algorithm and the scenario, the performance limitations of a single algorithm in different scenarios are avoided, and a balance between overall scheduling efficiency and quality is achieved, providing technical support for the economical and efficient operation of self-service car wash machines.

[0035] In a preferred embodiment, this application can be further configured such that, after the step of adjusting the power supply parameters and order reservation strategy of the self-service car wash based on the power supply scheduling strategy, the following steps are performed:

[0036] Real-time performance monitoring of self-service car wash machines after adjustments based on power supply scheduling strategies;

[0037] The power supply dispatch strategy is calibrated based on real-time performance monitoring results.

[0038] By adopting the above technical solution, this application introduces a dual feedback mechanism of real-time performance monitoring and parameter calibration after the power supply scheduling strategy is executed, thereby achieving dynamic optimization and continuous improvement of the scheduling system. It establishes a complete closed-loop verification system by collecting real-time performance key indicators such as power consumption and order completion rate of the car wash machine during actual operation and comparing them with the expected goals of the scheduling strategy. When the actual operating effect deviates from expectations, the parameter calibration program is automatically triggered to dynamically adjust core parameters such as the weight coefficients of the prediction model and the constraints of the optimization algorithm. This closed-loop control mechanism enables the power supply strategy to be continuously corrected according to the actual operating status of the equipment, overcoming the problem of strategy failure caused by sudden environmental changes or demand fluctuations, and significantly improving the system's adaptability to complex operating scenarios.

[0039] The second objective of this invention is achieved through the following technical solution:

[0040] A power supply dispatching system for a self-service car wash machine includes:

[0041] The data acquisition module is used to acquire electricity consumption data and order data of the self-service car wash machines in the target network.

[0042] The data input module is used to input electricity consumption data and order data into the pre-trained prediction model, so that the prediction model can output demand prediction information.

[0043] The strategy generation module is used to obtain power supply load information and electricity price fluctuation information, and generate power supply scheduling strategies based on the power supply load information, electricity price fluctuation information and demand forecast information.

[0044] The adjustment module is used to adjust the power supply parameters and order reservation strategy of the self-service car wash machine based on the power supply scheduling strategy.

[0045] By adopting the above technical solution, the data acquisition module is used to acquire electricity consumption data and order data of the self-service car wash machines in the target network; the data input module is used to input the electricity consumption data and order data into the pre-trained prediction model, so that the prediction model outputs demand prediction information; the strategy generation module is used to acquire power supply load information and electricity price fluctuation information, and generate power supply scheduling strategy based on the power supply load information, electricity price fluctuation information, and demand prediction information; the adjustment module is used to adjust the power supply parameters and order reservation strategy of the self-service car wash machines based on the power supply scheduling strategy.

[0046] In a preferred embodiment, this application may be further configured to include:

[0047] The environmental data acquisition layer module is used to acquire the current external environmental data and current time data of the target site;

[0048] The feature extraction layer module is used to extract environmental features, time features, electricity consumption features, and order features based on external environment data, current time data, electricity consumption data, and order data.

[0049] The time series analysis layer module is used to perform time series analysis based on environmental characteristics, time characteristics, electricity consumption characteristics, and order characteristics, thereby generating time series features including electricity demand and order demand.

[0050] The prediction output layer module is used to predict electricity demand based on time series features and output demand prediction information.

[0051] By adopting the above technical solution, the environmental data acquisition layer module is used to acquire the current external environment data and current time data of the target network point; the feature extraction layer module is used to extract environmental features, time features, electricity consumption features, and order features based on external environment data, current time data, electricity consumption data information, and order data information; the time series analysis layer module is used to perform time series analysis based on environmental features, time features, electricity consumption features, and order features, thereby generating time series features including electricity demand and order demand; and the prediction output layer module is used to predict electricity demand based on time series features and output demand prediction information.

[0052] In summary, this application includes at least one of the following beneficial technical effects:

[0053] 1. This application acquires electricity consumption and order data from self-service car wash machines within the target network and inputs it into a pre-trained prediction model to achieve accurate prediction of future electricity demand and order volume. By combining power load information and real-time electricity price fluctuation data, a dynamically optimized power supply scheduling strategy is constructed. Finally, based on this strategy, the equipment power supply parameters and order reservation schemes are adjusted synchronously to form a closed-loop control system. By integrating environmental factors and time series analysis into the prediction model, the shortcomings of fixed timetable mode in responding to environmental changes are overcome, enabling the power supply scheme to be dynamically adjusted according to external factors such as weather and holidays, thus avoiding the waste of electricity resources while ensuring service capacity. Secondly, by introducing electricity price fluctuation information as a core optimization parameter, a refined response to time-of-use pricing policies is achieved. By intelligently reducing power during high-price periods and improving service capacity during low-price periods, operating costs are significantly reduced. At the same time, power supply control and order management are optimized in synergy. When the peak grid load is predicted, the system will adjust the reservation strategy in advance, guiding users to use the service during off-peak hours through price leverage, which has the effect of ensuring the safe and stable operation of the power grid and maintaining a good user experience.

[0054] 2. This application achieves perception of external influencing factors through an environmental data acquisition layer, enabling the prediction results to respond promptly to special circumstances such as sudden weather changes; the feature extraction layer mines deep-seated correlation patterns from the raw data; the time series analysis layer performs time series analysis on multi-dimensional features to capture the changing patterns of electricity load and order volume; the prediction output layer finally transforms the time series analysis results into specific demand forecast values; the collaborative design of the above module layers improves the professionalism and accuracy of data processing at each stage, providing a reliable basis for demand forecasting for power supply dispatch.

[0055] 3. This application achieves economical and optimized scheduling under the condition of satisfying the power grid load constraints by intelligently dividing the electricity price sensitive interval and accurately solving the optimization algorithm. It avoids the subjectivity and lag of manually formulating scheduling strategies, and can dynamically adjust the power supply scheme according to electricity price fluctuations and demand changes, thus improving the accuracy and economy of power supply scheduling for self-service car wash machines.

[0056] 4. This application employs a dual-mode optimization strategy combining a mixed-integer programming algorithm and an improved greedy algorithm to achieve differentiated scheduling for different electricity price-sensitive intervals: For highly sensitive intervals, a mixed-integer programming algorithm is used for precise modeling and solving, treating equipment start-up and shutdown states as discrete variables and power parameters as continuous variables. Under the premise of satisfying grid load constraints, the algorithm obtains highly sensitive power supply parameters that balance economy and reliability. For low-sensitive intervals, an improved greedy algorithm is used for rapid optimization. This algorithm, through the principles of electricity price ranking and capacity priority allocation, obtains an approximate optimal solution with lower computational cost while ensuring basic power supply needs. The above optimization method ensures the accuracy of scheduling during highly sensitive periods and improves the computational efficiency during low-sensitive periods, enabling the rapid generation of optimal power supply schemes in complex and ever-changing operating environments. Through precise matching of the algorithm with the scenario, the performance limitations of a single algorithm in different scenarios are avoided, and a balance between overall scheduling efficiency and quality is achieved, providing technical support for the economical and efficient operation of self-service car wash machines.

[0057] 5. This application achieves dynamic optimization and continuous improvement of the scheduling system by introducing a dual feedback mechanism of real-time performance monitoring and parameter calibration after the power supply scheduling strategy is executed. It establishes a complete closed-loop verification system by collecting real-time performance key indicators such as power consumption and order completion rate of the car wash machine during actual operation and comparing them with the expected goals of the scheduling strategy. When the actual operating effect deviates from expectations, the parameter calibration program is automatically triggered to dynamically adjust core parameters such as the weight coefficients of the prediction model and the constraints of the optimization algorithm. This closed-loop control mechanism enables the power supply strategy to be continuously corrected according to the actual operating status of the equipment, overcoming the problem of strategy failure caused by sudden environmental changes or demand fluctuations, and significantly improving the system's adaptability to complex operating scenarios. Attached Figure Description

[0058] Figure 1 This is a flowchart of an embodiment of a power supply scheduling method for a self-service car wash machine according to this application;

[0059] Figure 2 This is a flowchart of step S20 in an embodiment of the power supply scheduling method for a self-service car wash machine according to this application;

[0060] Figure 3This is a flowchart of step S23 in an embodiment of the power supply scheduling method for a self-service car wash machine according to this application;

[0061] Figure 4 This is a flowchart illustrating step S30 in an embodiment of the power supply scheduling method for a self-service car wash machine according to this application. Detailed Implementation

[0062] The following is in conjunction with the appendix Figures 1-4 This application will be described in further detail.

[0063] In one embodiment, such as Figure 1 As shown, this application discloses a power supply scheduling method for a self-service car wash machine, which specifically includes the following steps:

[0064] S10: Obtain power consumption data and order data of the self-service car wash machines in the target network location;

[0065] In this embodiment, the target site is the specific service station where the self-service car wash machine is deployed, including the equipment cluster and its surrounding environment; the power consumption data information is the power parameters of the self-service car wash machine during operation, including real-time power, cumulative power consumption, voltage / current waveform, etc.; the order data information is the service record of user reservation and actual use, including time period, service type, completion time, etc.

[0066] Specifically, the system collects real-time electricity consumption data (such as smart meter readings) and order system data (such as reservation platform records) from the autonomous car wash machines at the target locations through IoT terminals, forming a structured dataset.

[0067] S20: Input the electricity consumption data and order data into the pre-trained prediction model, so that the prediction model outputs demand prediction information;

[0068] In this embodiment, the prediction model is a machine learning model trained on historical data, used to predict electricity demand and order volume in future periods; the demand prediction information is the structured prediction results output by the prediction model, including electricity load curves and order distribution predictions, etc.

[0069] Specifically, electricity consumption data and order data are input into a pre-trained prediction model (such as an LSTM neural network), which then analyzes historical patterns and real-time characteristics to output predictions of electricity demand and order volume for a future time period.

[0070] S30: Obtain power supply load information and electricity price fluctuation information, and generate power supply dispatching strategies based on power supply load information, electricity price fluctuation information and demand forecast information;

[0071] In this embodiment, the power supply load information is a composite dataset, including the maximum allowable load and real-time load rate that the power grid can currently provide, as well as the power supply load information of the car wash machine. The power supply load information of the car wash machine includes the rated power of a single machine, the total load of the cluster, the start-stop status of the equipment, and dynamic power consumption characteristics. The electricity price fluctuation information is a dynamic price table under the time-of-use electricity pricing policy, including the unit price and discount rules for each time period. The power supply scheduling strategy is a generated optimized control scheme, including the equipment power setpoint and order control rules.

[0072] Specifically, by combining the real-time load capacity and time-of-use electricity price provided by the power grid, the power supply load information of the self-service car wash machine, and the prediction results, the optimal scheduling scheme is calculated using optimization algorithms (such as mixed integer programming).

[0073] S40: Adjust the power supply parameters and order reservation strategy of the self-service car wash machine based on the power supply scheduling strategy.

[0074] In this embodiment, the order reservation strategy is a dynamically adjusted reservation rule, such as time-limited access and differentiated pricing;

[0075] Specifically, the generated power supply scheduling strategy is transformed into specific control instructions, including power supply control and order regulation.

[0076] In one embodiment, the prediction model includes an environmental data acquisition layer, a feature extraction layer, a time series analysis layer, and a prediction output layer, such as... Figure 2 As shown, step S20 includes the following steps:

[0077] S21: The environmental data acquisition layer acquires the current external environmental data and current time data of the target site;

[0078] S22: The feature extraction layer extracts environmental features, time features, electricity consumption features, and order features based on external environment data, current time data, electricity consumption data, and order data.

[0079] S23: The time series analysis layer performs time series analysis based on environmental characteristics, time characteristics, electricity consumption characteristics, and order characteristics, thereby generating time series characteristics including electricity demand and order demand;

[0080] S24: The prediction output layer predicts electricity demand based on time series features and outputs demand prediction information.

[0081] In this embodiment, the environmental data acquisition layer is the first layer of the prediction model, used to collect external environmental parameters of the target site and current time data. The external environmental parameters include meteorological data (temperature, precipitation probability, wind speed, etc.) and geographical data (traffic flow index within a preset range around the site). The time data includes the current time, weekday type (weekday / weekend), holiday markings, etc. The feature extraction layer is the second layer of the prediction model, used to convert the raw data into machine-recognizable features. The time series analysis layer is the third layer of the prediction model, used to analyze the changing patterns of features over time. The prediction output layer is the fourth layer of the prediction model, used to generate the final prediction result.

[0082] Specifically, the environmental data acquisition layer collects real-time external environmental data such as weather conditions and temperature at target sites, as well as specific time information, providing basic data support for prediction; the feature extraction layer extracts predictive feature vectors from environmental data, time information, historical electricity consumption records, and order data, including environmental features reflecting the impact of weather, time features reflecting periodic changes, electricity consumption features characterizing equipment operating status, and order features showing user behavior patterns; the time series analysis layer performs time series analysis on environmental features, time features, electricity consumption features, and order features to generate time series features including electricity demand and order demand; and the prediction output layer predicts electricity demand based on the time series features and outputs demand prediction information.

[0083] For example, in the scenario of a car wash station in a commercial area during a spring weekday, the process of acquiring environmental data and current time data is as follows: Current weather: light rain (70% probability of precipitation), time stamp: Thursday morning 10:15 (non-holiday); Feature extraction process: generating a "rainy day effect coefficient" of 1.8 (historical data shows that orders on rainy days increase by 80%), and marking it as "off-peak period on weekday mornings" (baseline load 12kW); Time series analysis process: identifying the unique pattern of this station: rainy day orders are concentrated during the lunch break (12:00-13:00) and peak electricity consumption occurs 2 hours after rain (evaporation of residual water stains requires enhanced drying); Prediction output process: generating differentiated predictions: 10:00-11:00 → 2 orders (normal), 12:00-13:00 → 8 orders (surge on rainy days), 14:00-15:00 → 25kW power supply capacity needs to be reserved (drying demand).

[0084] In one embodiment, such as Figure 3 As shown, step S23 includes the following steps:

[0085] S231: The time series analysis layer performs weighted fusion of environmental features, time features, electricity consumption features, and order features to generate a joint feature vector;

[0086] S232: The time series analysis layer performs a periodic-based trend decomposition on the joint feature vector to obtain the long-term trend component and the seasonal periodic component.

[0087] S233: The time series analysis layer performs dynamic modeling based on neural networks on the long-term trend component and the seasonal cycle component to obtain the baseline forecast of electricity load and the order fluctuation pattern.

[0088] S234: The time series analysis layer generates time series features based on baseline electricity load forecasts and order fluctuation patterns.

[0089] In this embodiment, weighted fusion is a calculation process that uses an attention mechanism to dynamically assign weights to multi-dimensional features (environmental features, time features, electricity consumption features, and order features). A trainable parameter matrix is ​​used to differentiate the importance of different feature vectors. The joint feature vector is a unified feature representation formed after weighted fusion, whose dimension is equal to a weighted linear combination of the input feature dimensions, serving as input for subsequent time series analysis. Trend decomposition is an algorithmic process that performs Seasonal Trend Decomposition (STL) on the joint feature vector, decomposing the time series data into three components: a long-term trend component (reflecting slow changes such as equipment aging and user growth), a seasonal cycle component (reflecting...), and a seasonal periodic component (reflecting...). The system employs a fixed weekly / daily pattern (using Fourier basis functions for fitting) and residual components (residual random fluctuation terms); dynamic modeling utilizes a neural network architecture with time-series processing capabilities to establish prediction models for each decomposed component; the baseline electricity load forecast is a basic electricity demand curve output by a trend prediction network, reflecting the equipment baseline energy consumption without considering periodic fluctuations; the order fluctuation pattern is a periodic demand change pattern generated by a periodic prediction network, quantified as an order volume fluctuation coefficient matrix at different time scales; the time series features are the complete prediction results reconstructed by superimposing the baseline forecast and the fluctuation pattern according to the superposition principle, including time-series forecasts of electricity demand and order volume with confidence intervals;

[0090] Specifically, by using feature weighted fusion, environmental features, time features, electricity consumption features, and order features are integrated into a unified joint feature vector to ensure that various influencing factors are reasonably balanced. The joint feature vector is decomposed into a trend component reflecting long-term changes and a seasonal component reflecting cyclical patterns, so that demand features at different time scales can be processed separately. For the decomposed components, neural networks are used for dynamic modeling, where the long-term trend component is used to predict the basic electricity load, and the seasonal cyclical component is used to capture the fluctuation pattern of order volume. Finally, the prediction results of the two components are recombined to form a complete time series feature.

[0091] For example, taking the winter operation forecast of a community car wash station as a scenario, the data input process is as follows: Environment: snowy day (temperature -5℃), Time: December 24 (Sunday + Christmas Eve), Electricity consumption: recent average load 14kW, Orders: 12 orders on the same day last week; The weighted fusion process generates a joint vector: [0.9 (severe cold), 1.5 (holiday), 0.7 (electricity consumption), 1.3 (orders)]; The trend decomposition process is as follows: Trend item: average daily load in winter 15.2kW (13.8kW in November), Periodic item: Sunday order volume is usually 40% higher than midweek; The dynamic modeling process is as follows: Trend network: predicts that the load will reach 16kW during Christmas week, Periodic network: identifies an additional 20% increase in orders on Christmas Eve; The output process is as follows: December 24 forecast: Electricity load: 16×(1+0.2)=19.2kW, Order volume: 12×1.4×1.2≈20 orders.

[0092] In one embodiment, such as Figure 4 As shown, step S30 includes the following steps:

[0093] S31: Construct a timeline mapping table for electricity price fluctuation information and divide the electricity price sensitive range based on demand forecast information;

[0094] S32: Solve for the optimal power supply parameters based on a pre-set optimization algorithm combined with electricity price sensitive range and power supply load information;

[0095] S33: Generate power supply scheduling strategy based on optimal power supply parameters.

[0096] In this embodiment, the time axis mapping table for electricity price fluctuation information is a structured table that establishes a precise mapping relationship between time-of-use electricity price data and time nodes. It contains the electricity price value and its change gradient corresponding to each time slot (usually a 15-minute interval). The time axis mapping table for electricity price fluctuation information supports quick querying of the electricity price status at any time. The electricity price sensitive interval is a time period classification based on demand forecasting and electricity price fluctuation characteristics. The optimization solution algorithm is a mathematical programming method used to calculate the optimal power supply parameters, such as mixed integer programming (handling discrete decision variables such as equipment start-up and shutdown) and improved greedy algorithm (achieving fast approximate solution). The optimal power supply parameters are a set of equipment control parameters obtained through optimization calculation, including: target power values ​​for each time period, equipment start-up and shutdown sequence arrangement, power adjustment gradient limit, etc.

[0097] Specifically, time-of-use electricity price data is structured and mapped along a time axis to establish a precise correspondence between electricity prices and time points. Based on demand forecasting information, different time periods are divided into high-sensitivity and low-sensitivity intervals, with high-sensitivity intervals corresponding to periods with higher electricity prices and greater predicted demand, and low-sensitivity intervals the opposite. On this basis, a preset optimization algorithm is used to comprehensively consider the results of the electricity price-sensitive interval division and real-time power supply load information. Through mathematical modeling and constraint solving, the optimal power supply parameters are obtained, including the optimal power settings and equipment operation schemes for each time period. An executable power dispatching strategy is generated based on the optimal power supply parameters obtained from the solution.

[0098] In one embodiment, the electricity price sensitive range includes a high-sensitivity range and a low-sensitivity range, and the optimal power supply parameters include high-sensitivity power supply parameters and low-sensitivity power supply parameters. Step S32 includes the following steps:

[0099] S321: Solve the high-sensitivity power supply parameters by combining a pre-set mixed integer programming algorithm with information on the high-sensitivity interval and power supply load.

[0100] S322: Based on a pre-set improved greedy algorithm combined with low-sensitivity intervals and power supply load information, the low-sensitivity power supply parameters are obtained.

[0101] In this embodiment, the high-sensitivity interval is the period when the electricity price is higher than the threshold and the predicted demand exceeds the preset capacity of the equipment; the low-sensitivity interval is the period when the electricity price is lower than the average and the predicted demand is less than the preset capacity; mixed integer programming is an optimization method for handling discrete decision variables such as equipment start-up and shutdown; the improved greedy algorithm is an optimization method for achieving fast approximate solution, which adds equipment load balancing factors, electricity price change trend prediction, and power regulation cost weights to the standard greedy algorithm; the high-sensitivity power supply parameters are precise control parameters calculated by mixed integer programming, including the optimal start-up and shutdown combination of equipment (binary decision variables), the power setpoint of each equipment (continuous variables), and the minimum switching time constraint; the low-sensitivity power supply parameters are approximately optimal parameters generated based on the improved greedy algorithm, including the basic power allocation scheme, the equipment rotation sequence, and the flexible power adjustment range;

[0102] Furthermore, a power supply scheduling strategy is generated based on the obtained highly sensitive power supply parameters and low sensitive power supply parameters. The order reservation strategy in this power supply scheduling strategy includes dynamic time-of-use pricing strategy, demand diversion strategy, weather elasticity strategy, etc.

[0103] The dynamic time-based pricing strategy involves dynamic pricing based on both high-sensitivity and low-sensitivity power supply zones. The demand diversion strategy guides demand across these zones, including capacity limits (e.g., hiding some reservation time slots), tiered rewards, and intelligent time slot recommendations for high-sensitivity zones, and capacity release (e.g., adding discounts for off-peak hours), basic rewards, and bulk reservation incentives for low-sensitivity zones. The weather-flexible strategy combines settings for both high-sensitivity and low-sensitivity power supply zones with consideration of extreme weather conditions. For example, if the reservation cancellation rate exceeds a preset value during heavy rain, reservations for the next few hours are automatically closed based on weather forecasts, and the released power capacity is converted to energy storage charging. Additionally, nighttime discount periods are opened during continuous heat dissipation scenarios.

[0104] Specifically, for the highly sensitive range, a mixed-integer programming algorithm is used for accurate modeling and solving. The start-up and shutdown status of the equipment is treated as discrete variables and the power parameters as continuous variables. Under the premise of satisfying the grid load constraints, the highly sensitive power supply parameters that take into account both economy and reliability are obtained. For the low-sensitivity range, an improved greedy algorithm is used for fast optimization. This algorithm obtains an approximate optimal solution with low computational cost while ensuring basic power supply needs through the principles of electricity price ranking and capacity priority allocation.

[0105] In one embodiment, after step S40, the following steps are performed:

[0106] S41: Real-time performance monitoring of self-service car wash machines after adjustments based on power supply scheduling strategies;

[0107] S42: Perform parameter calibration on power supply dispatching strategy based on real-time performance monitoring results.

[0108] In this embodiment, real-time performance monitoring refers to the dataset of car wash machine operating status collected in real time by IoT sensors, including actual power consumption (compared to the scheduling target value), order completion time (compared to the reservation time), equipment response delay (time difference between instruction issuance and execution), and grid load fluctuation (deviation rate between actual and predicted values). Parameter calibration is an optimization process that dynamically adjusts strategy parameters based on monitoring data. The main calibration objects include the error compensation coefficient of the prediction model, the weight parameters of the optimization algorithm (such as the start-stop penalty coefficient M), the threshold for dividing the electricity price sensitive area, and the buffer coefficient of the grid safety margin.

[0109] Specifically, by collecting real-time performance key indicators such as power consumption and order completion rate of the car wash machine during actual operation, and comparing and analyzing them with the expected goals of the scheduling strategy, a complete closed-loop verification system is established; when the actual operation effect is detected to deviate from the expectation, the parameter calibration program is automatically triggered to dynamically adjust the core parameters such as the weight coefficient of the prediction model and the constraints of the optimization algorithm.

[0110] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0111] In one embodiment, a power supply scheduling system for a self-service car wash is provided, which corresponds one-to-one with the power supply scheduling method for a self-service car wash described in the above embodiment. The power supply scheduling system for the self-service car wash includes:

[0112] The data acquisition module is used to acquire electricity consumption data and order data of the self-service car wash machines in the target network.

[0113] The data input module is used to input electricity consumption data and order data into the pre-trained prediction model, so that the prediction model can output demand prediction information.

[0114] The strategy generation module is used to obtain power supply load information and electricity price fluctuation information, and generate power supply scheduling strategies based on the power supply load information, electricity price fluctuation information and demand forecast information.

[0115] The adjustment module is used to adjust the power supply parameters and order reservation strategy of the self-service car wash machine based on the power supply scheduling strategy;

[0116] Optional, also includes:

[0117] The environmental data acquisition layer module is used to acquire the current external environmental data and current time data of the target site;

[0118] The feature extraction layer module is used to extract environmental features, time features, electricity consumption features, and order features based on external environment data, current time data, electricity consumption data, and order data.

[0119] The time series analysis layer module is used to perform time series analysis based on environmental characteristics, time characteristics, electricity consumption characteristics, and order characteristics, thereby generating time series features including electricity demand and order demand.

[0120] The prediction output layer module is used to predict electricity demand based on time series features and output demand prediction information.

[0121] Specific limitations regarding the power supply scheduling system for a self-service car wash machine can be found in the above description of the power supply scheduling method for a self-service car wash machine, and will not be repeated here. Each module in the aforementioned power supply scheduling system for a self-service car wash machine can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0122] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A power supply scheduling method for a self-service car wash machine, characterized in that: Including the following steps: Obtain electricity consumption and order data for self-service car wash machines within the target network. Electricity consumption data and order data are input into a pre-trained prediction model. The prediction model includes an environmental data acquisition layer, a feature extraction layer, a time series analysis layer, and a prediction output layer, enabling the model to output demand prediction information. The steps include: The environmental data acquisition layer acquires the current external environmental data and current time data of the target site; The feature extraction layer extracts environmental features, time features, electricity consumption features, and order features based on external environment data, current time data, electricity consumption data, and order data. The time series analysis layer performs time series analysis based on environmental features, time features, electricity consumption features, and order features to generate time series features including electricity demand and order demand. This includes: weighted fusion of environmental features, time features, electricity consumption features, and order features to generate a joint feature vector; periodic trend decomposition of the joint feature vector to obtain long-term trend components and seasonal periodic components; dynamic modeling of the long-term trend components and seasonal periodic components based on neural networks to obtain electricity load baseline predictions and order fluctuation patterns; and the generation of time series features based on electricity load baseline predictions and order fluctuation patterns. The prediction output layer predicts electricity demand based on time series features and outputs demand prediction information. Obtain power supply load information and electricity price fluctuation information, and generate a power supply dispatch strategy based on the power supply load information, electricity price fluctuation information, and demand forecast information, including the following steps: Construct a time-axis mapping table for electricity price fluctuation information, and divide the electricity price sensitive intervals based on demand forecast information. The electricity price sensitive intervals include high-sensitivity intervals and low-sensitivity intervals. The optimal power supply parameters are obtained by combining a pre-set optimization algorithm with electricity price sensitive intervals and power supply load information. The optimal power supply parameters include highly sensitive power supply parameters and low-sensitive power supply parameters. Specifically, the high-sensitive power supply parameters are obtained by combining a pre-set mixed integer programming algorithm with high-sensitive intervals and power supply load information, and the low-sensitive power supply parameters are obtained by combining a pre-set improved greedy algorithm with low-sensitive intervals and power supply load information. Generate power supply scheduling strategies based on optimal power supply parameters; The power supply parameters and order reservation strategies for self-service car wash machines are adjusted based on power supply scheduling strategies.

2. The power supply scheduling method for a self-service car wash machine according to claim 1, characterized in that: After the steps of adjusting the power supply parameters and order reservation strategy of the self-service car wash machine based on the power supply scheduling strategy, the following steps are performed: Real-time performance monitoring of self-service car wash machines after adjustments based on power supply scheduling strategies; The power supply dispatch strategy is calibrated based on real-time performance monitoring results.

3. A power supply scheduling system for a self-service car wash machine, used to implement the power supply scheduling method for a self-service car wash machine as described in claim 1, characterized in that: include: The data acquisition module is used to acquire electricity consumption data and order data of the self-service car wash machines in the target network. The data input module is used to input electricity consumption data and order data into the pre-trained prediction model, so that the prediction model can output demand prediction information. The strategy generation module is used to obtain power supply load information and electricity price fluctuation information, and generate power supply scheduling strategies based on the power supply load information, electricity price fluctuation information and demand forecast information. The adjustment module is used to adjust the power supply parameters and order reservation strategy of the self-service car wash machine based on the power supply scheduling strategy; The environmental data acquisition layer module is used to acquire the current external environmental data and current time data of the target site; The feature extraction layer module is used to extract environmental features, time features, electricity consumption features, and order features based on external environment data, current time data, electricity consumption data, and order data. The time series analysis layer module is used to perform time series analysis based on environmental characteristics, time characteristics, electricity consumption characteristics, and order characteristics, thereby generating time series features including electricity demand and order demand. The prediction output layer module is used to predict electricity demand based on time series features and output demand prediction information.

Images