User adjustable resource regulation and control method and system based on machine learning and medium

Abstract

The invention discloses a user adjustable resource regulation and control method and system based on machine learning, and a medium, mainly relates to the technical field of resource regulation and control, and is used for solving the problem that an existing scheme mainly pays attention to the total electricity consumption of a user main body. The problems that the time sequence characteristics of power utilization of adjustable resource equipment and the influence of environmental influence factors on the power utilization behavior of the equipment are ignored, the scheme for evaluating the adjustment potential of the adjustable resources on the user side is simple, and a regulation and control method cannot meet the personalized requirements of users are solved. Comprising the steps of obtaining a specific adjustment value of an adjustment potential coefficient, and obtaining a device ID meeting a specific actual value and the specific adjustment value; inputting the equipment ID meeting the specific actual value and the specific adjustment value into a trained second machine learning algorithm to obtain predicted time sequence electricity consumption corresponding to the equipment ID in a future analysis interval; and when the sum power consumption of all the predicted time sequence power consumption is greater than the difference value between the predicted total power consumption and the power consumption regulation and control peak value, issuing a regulation and control instruction to a user side corresponding to the equipment ID.

Description

Technical Field

[0001] The present application relates to the technical field of adjustable resource regulation, and in particular to a user-adjustable resource regulation method, system and medium based on machine learning. Background Art

[0002] With the continuous growth of energy demand and the increasing complexity of the power system, the role of user-side adjustable resources in balancing power supply and demand and in the stable operation of the power grid is becoming increasingly important. User-side adjustable resource equipment, such as smart home appliances, distributed energy storage devices, adjustable lighting systems, etc., can effectively alleviate the power supply pressure during peak hours of the power grid and improve the overall efficiency and reliability of the power system by flexibly adjusting their power consumption behavior. However, how to accurately and efficiently regulate these user-side adjustable resources to achieve the optimal allocation of power resources has become a key issue that needs to be urgently addressed in the current energy management and smart grid fields.

[0003] Existing methods mainly focus on the total electricity consumption of the user entity, ignoring the timing characteristics of the electricity consumption of adjustable resource equipment and the impact of environmental factors on the electricity consumption behavior of the equipment. For example, the electricity consumption of equipment such as air conditioners and electric heaters will vary significantly under different seasons and weather conditions, and existing methods are difficult to fully consider these factors, resulting in inaccurate prediction and regulation of equipment electricity consumption behavior. In addition, when evaluating the regulation potential of adjustable resources on the user side, existing methods usually use simple fixed coefficients or empirical values, and fail to fully consider multiple factors such as the actual operating conditions of the equipment, environmental factors, and user needs. In addition, different users have different electricity usage habits and equipment configurations, and existing regulation methods often use a unified regulation strategy that cannot meet the personalized needs of users. Summary of the invention

[0004] The present application provides a user-adjustable resource control method, system and medium based on machine learning to solve the problems that the existing solutions mainly focus on the total power consumption of the user entity, ignore the timing characteristics of the power consumption of the adjustable resource equipment and the impact of environmental factors on the power consumption behavior of the equipment, the solution for evaluating the adjustment potential of the user-side adjustable resources is simple, and the control method cannot meet the personalized needs of users.

[0005] In a first aspect, the present application provides a user-adjustable resource control method based on machine learning, the method comprising: Collect the power consumption impact data of the user-side adjustable resource equipment every hour within the preset time period; the power consumption impact data includes: the equipment time series power consumption and environmental impact factors corresponding to the adjustable resource equipment, and the adjustable resource equipment contains a unique equipment ID; Annotate the regulation potential coefficient corresponding to the hourly electricity consumption impact data within a preset time period, and then train the first machine learning algorithm using the collection time, device ID, environmental impact factors, and regulation potential system to obtain a trained first machine learning algorithm; Using the collection time, the device ID, and the device sequential power consumption, the second machine learning algorithm is trained to obtain a trained second machine learning algorithm; Collecting household time-series power consumption at the user side of a preset control interval within a preset time period, and then using the collection time and household time-series power consumption to train a third machine learning algorithm to obtain a trained third machine learning algorithm; Obtaining the peak value of power consumption regulation in the future analysis interval, and obtaining the predicted total power consumption of the preset regulation interval in the future analysis interval through the trained third machine learning algorithm; When the predicted total power consumption is greater than or equal to the power consumption regulation peak value, the specific actual value of the environmental influencing factor is obtained, and the specific adjustment value of the regulation potential coefficient is obtained through the preset adjustment rule, and the trained first machine learning algorithm is input to obtain the device ID that meets the specific actual value and the specific adjustment value; The device ID that meets the specific actual value and the specific adjustment value is input into the trained second machine learning algorithm to obtain the predicted time series power consumption corresponding to the device ID in the future analysis interval; When the sum of all predicted time-series power consumption is greater than the difference between the predicted total power consumption and the power consumption control peak value, a control instruction is sent to the user side corresponding to the device ID to obtain the device ID that returns the executable instruction; When the summed power consumption corresponding to the device ID that returns the executable instruction is less than the difference between the predicted total power consumption and the power consumption regulation peak value or a further downward adjustment instruction is received, the specific adjustment value of the adjustment potential coefficient is adjusted based on the preset adjustment rules.

[0006] In one implementation of the present application, before collecting the power consumption impact data of the user-side adjustable resource device every hour within a preset time period, the method further includes: Obtain the original adjustable resource equipment information uploaded by the user through the preset data collection interface; The original adjustable resource equipment information includes at least: specific model information and rated power of the adjustable resource equipment; Determine whether the specific model information and rated power meet the preset limiting rules corresponding to the adjustable resource equipment; When the preset limiting rule is met, determining that the device corresponding to the original adjustable resource device information is an adjustable resource device; A unique device ID is randomly generated for the adjustable resource device and stored.

[0007] In one implementation of the present application, after obtaining the peak value of power consumption regulation in the future analysis interval and obtaining the predicted total power consumption of the preset regulation interval in the future analysis interval through the trained third machine learning algorithm, the method further includes: When the predicted total power consumption is less than or equal to the power consumption control peak value, the default program is maintained.

[0008] In one implementation of the present application, before collecting the power consumption impact data of the user-side adjustable resource device every hour within a preset time period, the method further includes: Configure smart meter terminals, environment sensing terminals and edge computing gateways for adjustable resource devices; Then, the smart meter terminal is used to collect the equipment's time-series power consumption and upload it to the edge computing gateway; Use environmental sensing terminals to collect environmental impact factors and upload them to the edge computing gateway; Through the edge computing gateway, data alignment is performed on the device’s sequential power consumption and environmental influencing factors.

[0009] In one implementation of the present application, a control instruction is sent to the user side corresponding to the device ID to obtain the device ID that returns the executable instruction, specifically including: Send control instructions to the user side corresponding to the device ID, and obtain execution instructions returned by the user side corresponding to the device ID; wherein the execution instructions are divided into executable and non-executable; Determines the device ID for which the executable command is returned.

[0010] In one implementation of the present application, the preset adjustment rules specifically include: The specific adjustment value of the potential coefficient is adjusted in a manner of successively decreasing the preset value; wherein the preset adjustment rule includes a preset initial adjustment value.

[0011] In one implementation of the present application, the summed power consumption corresponding to the device ID of the returned executable instruction is less than the difference between the predicted total power consumption and the power consumption regulation peak value, or a further downward adjustment instruction is received, specifically including: Get the predicted time series power consumption corresponding to the device ID that returns the executable instruction; Accumulate the predicted time series power consumption to obtain the summed power consumption; Calculate whether the total power consumption is less than the difference between the predicted total power consumption and the power consumption control peak value; Obtain further downward adjustment instructions through the preset information acquisition interface.

[0012] In a second aspect, the present application provides a user-adjustable resource control method system based on machine learning, the system comprising: A collection module is used to collect the power consumption impact data of the user-side adjustable resource equipment every hour within a preset time period; wherein the power consumption impact data includes: the equipment time sequence power consumption and environmental impact factors corresponding to the adjustable resource equipment, and the adjustable resource equipment includes a unique equipment ID; A training module is used to mark the adjustment potential coefficient corresponding to the hourly electricity consumption impact data within a preset time period, and then use the collection time, device ID, environmental impact factors and adjustment potential system to train the first machine learning algorithm to obtain a trained first machine learning algorithm; use the collection time, device ID, and device time-series power consumption to train the second machine learning algorithm to obtain a trained second machine learning algorithm; collect the household time-series power consumption on the user side of the preset regulation interval within the preset time period, and then use the collection time and household time-series power consumption to train the third machine learning algorithm to obtain a trained third machine learning algorithm; An acquisition module is used to obtain a peak value of power consumption regulation within a future analysis interval, and to obtain a predicted total power consumption of a preset regulation interval within the future analysis interval through a trained third machine learning algorithm; when the predicted total power consumption is greater than or equal to the peak value of power consumption regulation, the specific actual value of the environmental influencing factor is obtained, and the specific adjustment value of the regulation potential coefficient is obtained through a preset adjustment rule, and the trained first machine learning algorithm is input to obtain a device ID that satisfies the specific actual value and the specific adjustment value; the device ID that satisfies the specific actual value and the specific adjustment value is input into a trained second machine learning algorithm to obtain the predicted time series power consumption corresponding to the device ID in the future analysis interval; when the sum of all predicted time series power consumption is greater than the difference between the predicted total power consumption and the power consumption regulation peak, a control instruction is issued to the user side corresponding to the device ID to obtain the device ID that returns an executable instruction; The adjustment module is used to adjust the specific adjustment value of the adjustment potential coefficient based on the preset adjustment rules when the summed power consumption corresponding to the device ID that returns the executable instruction is less than the difference between the predicted total power consumption and the power consumption regulation peak or when a further downward adjustment instruction is received.

[0013] In one implementation of the present application, the adjustment module includes an adjustment unit. Used to obtain the predicted time series power consumption corresponding to the device ID that returns the executable instruction; Accumulate the predicted time series power consumption to obtain the summed power consumption; Calculate whether the total power consumption is less than the difference between the predicted total power consumption and the power consumption control peak value; Obtain further downward adjustment instructions through the preset information acquisition interface.

[0014] In a third aspect, the present application provides a non-volatile computer storage medium having computer instructions stored thereon, which, when executed, implement a user-adjustable resource control method based on machine learning as any of the above.

[0015] It can be seen from the above technical solutions that this application has the following advantages: Existing solutions mainly focus on the total electricity consumption of the user entity, while this application collects the hourly electricity consumption impact data of the user-side adjustable resource equipment, especially the equipment's time-series electricity consumption, which makes the analysis more detailed and can capture the dynamic changes in electricity consumption behavior.

[0016] This application not only takes into account the power consumption of the equipment, but also incorporates environmental factors (such as temperature, humidity, light, etc.). These factors have an important impact on the power consumption behavior of the equipment. Including them in the analysis can more accurately evaluate the power consumption of the equipment and the regulation potential.

[0017] ‌This application establishes a personalized regulation potential evaluation system for each device by marking the regulation potential coefficient corresponding to the hourly electricity consumption impact data. This helps to more accurately identify which devices have greater regulation space, thereby formulating more effective regulation strategies.

[0018] The collected data is used to train the first machine learning algorithm, making the evaluation of the adjustment potential coefficient more intelligent and automated, thereby improving the accuracy and efficiency of the evaluation.

[0019] ‌The first machine learning algorithm‌: Combining the collection time, device ID, environmental influencing factors and adjustment potential coefficient, it trains the device ID that can predict specific environmental conditions and adjustment requirements, providing a basis for precise control.

[0020] ‌The second machine learning algorithm‌: Using the collection time, device ID and device time series power consumption, a model that can predict the predicted time series power consumption corresponding to the device ID is trained, which helps to more accurately estimate the power consumption after regulation.

[0021] ‌The third machine learning algorithm‌: By collecting household electricity consumption in real time, a model is trained that can predict the total electricity consumption in the preset control range within the future analysis interval, providing forward-looking guidance for electricity consumption regulation.

[0022] By comparing the predicted total power consumption and the power control peak, determine whether control measures need to be taken. When the predicted total power consumption is greater than or equal to the power control peak, the device ID that needs to be controlled is determined through the trained machine learning algorithm based on the environmental influencing factors and the specific adjustment value of the adjustment potential coefficient, and the control instruction is issued. According to the executable instructions and control effects returned by the device, the specific adjustment value of the adjustment potential coefficient is dynamically adjusted to achieve continuous optimization of the control strategy.

[0023] This application can better meet the personalized needs of users through refined electricity consumption data collection, personalized regulation potential assessment, multi-dimensional machine learning model construction, and dynamic regulation strategy formulation and execution. For example, for users with different electricity consumption habits and different environmental conditions, a regulation strategy that is more in line with their actual situation can be formulated to improve user satisfaction and participation.

[0024] In summary, the present application provides a user-adjustable resource regulation method, system and medium based on machine learning. Through refined electricity consumption data collection and analysis, personalized regulation potential assessment, multi-dimensional machine learning model construction, and dynamic regulation strategy formulation and execution, it effectively solves the problems of existing solutions that mainly focus on the total electricity consumption of the user entity, ignore the timing characteristics of the electricity consumption of adjustable resource equipment and the impact of environmental factors on the electricity consumption behavior of the equipment, the scheme for evaluating the regulation potential of user-side adjustable resources is simple, and the regulation method cannot meet the personalized needs of users. It improves the accuracy and efficiency of electricity consumption regulation and better meets the personalized needs of users. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solution of the present invention, the accompanying drawings required for use in the description will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0026] Figure 1 This is a flow chart of a user-adjustable resource control method based on machine learning provided in an embodiment of the present application.

[0027] Figure 2 It is a schematic diagram of the internal structure of a user-adjustable resource control method system based on machine learning provided in an embodiment of the present application. DETAILED DESCRIPTION

[0028] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0029] It should be understood by those skilled in the art that the embodiments described below are only preferred embodiments of the present disclosure, and do not mean that the present disclosure can only be implemented through the preferred embodiments. The preferred embodiments are only used to explain the technical principles of the present disclosure, and are not used to limit the protection scope of the present disclosure. Based on the preferred embodiments provided by the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work should still fall within the protection scope of the present disclosure.

[0030] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0031] The technical solution proposed in the embodiments of the present application is described in detail below with reference to the accompanying drawings.

[0032] The embodiment provides a user-adjustable resource control method based on machine learning, such as Figure 1 As shown, the method provided in the embodiment of the present application mainly includes the following steps: Step 110: Collect the power consumption impact data of the user-side adjustable resource equipment every hour within a preset time period.

[0033] The power consumption impact data includes: the equipment time-series power consumption and environmental impact factors corresponding to the adjustable resource equipment, and the adjustable resource equipment contains a unique equipment ID.

[0034] Those skilled in the art can understand that by collecting the power consumption impact data of the user-side adjustable resource equipment every hour within a preset time period, including the equipment's time-series power consumption and environmental impact factors, a more comprehensive understanding of the equipment's power consumption behavior and its impact by environmental factors can be achieved. This helps to more accurately evaluate the equipment's power consumption pattern and regulation potential, and provide data support for subsequent regulation strategies.

[0035] In some embodiments, before collecting the power consumption impact data of the user-side adjustable resource device every hour within the preset time period, the method further includes: The original adjustable resource device information uploaded by the user side is obtained through a preset data collection interface; wherein the original adjustable resource device information includes at least: specific model information and rated power of the adjustable resource device; determining whether the specific model information and rated power meet the preset limitation rules corresponding to the adjustable resource device; when the preset limitation rules are met, determining that the device corresponding to the original adjustable resource device information is an adjustable resource device; randomly generating a unique device ID for the adjustable resource device, and storing the device ID.

[0036] It should be noted that before collecting data, the original adjustable resource equipment information uploaded by the user is obtained through the preset data collection interface, and verified and screened to ensure that only equipment that meets the preset limit rules is included in the control range. This helps to ensure the accuracy and reliability of the data.

[0037] A unique device ID is randomly generated for each adjustable resource device to facilitate subsequent data processing, analysis, and the issuance of control instructions. This unique identification method simplifies the device management and tracking process and improves control efficiency.

[0038] Through the detailed collection of electricity consumption impact data and standardized processing of equipment information, it is possible to more accurately evaluate the electricity consumption and regulation potential of each device, thereby formulating more personalized regulation strategies.

[0039] The use of a unique device ID enables control instructions to be sent directly to specific devices, reducing intermediate links and improving the response speed and execution efficiency of control.

[0040] Specific examples: Assume that there is a smart home system, which contains a variety of adjustable resource devices, such as smart air conditioners, smart lighting, and smart sockets. In order to achieve precise control of these devices, the system adopts the above method.

[0041] The system obtains the original adjustable resource equipment information uploaded by the user through the preset data collection interface, including the specific model (such as "Model A") and rated power (such as "1.5 horsepower") of the smart air conditioner.

[0042] The system verifies whether this information meets the preset limiting rules (for example, model A is in the list of devices that support regulation and the rated power is within a reasonable range).

[0043] When the rules are met, the system randomly generates a unique device ID (such as "Device_001") for the smart air conditioner and stores the device ID.

[0044] Next, the system collects the power consumption impact data of the smart air conditioner every hour within a preset time period (such as a week), including the equipment's time-series power consumption (such as hourly power consumption) and environmental influencing factors (such as indoor and outdoor temperature, humidity, etc.).

[0045] Before collecting the power consumption impact data of the user-side adjustable resource equipment every hour within the preset time period, the method further includes: Configure smart meter terminals, environment sensing terminals and edge computing gateways for adjustable resource devices; Then, the smart meter terminal is used to collect the equipment's time-series power consumption and upload it to the edge computing gateway; Use environmental sensing terminals to collect environmental impact factors and upload them to the edge computing gateway; Through the edge computing gateway, data alignment is performed on the device’s sequential power consumption and environmental influencing factors.

[0046] Those skilled in the art can understand that, by configuring smart meter terminals and environment perception terminals for adjustable resource devices, the present application can collect the time-series power consumption and environmental influencing factors of the device in real time and accurately. The smart meter terminal focuses on the collection of power data, while the environment perception terminal is responsible for capturing environmental parameters. The combination of the two provides comprehensive data support.

[0047] The introduction of the edge computing gateway enables data from different terminals to be integrated and pre-processed locally. The data alignment function ensures the consistency of power data and environmental data in the time dimension, providing a reliable data foundation for subsequent analysis and decision-making.

[0048] Since the data is initially processed at the edge computing gateway, only the necessary data and analysis results are uploaded to the cloud or central server, which reduces network load and transmission delay. At the same time, the fast response capability of the edge computing gateway also improves the real-time performance of the system.

[0049] By configuring smart meter terminals, environmental sensing terminals and edge computing gateways, the system can flexibly adapt to the needs of different devices and scenarios. When new devices or functions need to be added, it is only necessary to simply add the corresponding terminals and configure the gateways without large-scale transformation of the entire system.

[0050] Specific examples: Assume that in an industrial park, there are multiple adjustable resource devices, such as motors, air conditioning units, etc. In order to achieve precise control and energy efficiency management of these devices, the above method is adopted.

[0051] A smart meter terminal is configured for each motor and air-conditioning unit to collect the equipment's sequential power consumption in real time.

[0052] At the same time, an environmental sensing terminal is configured to collect environmental influencing factors such as temperature, humidity, light intensity, etc.

[0053] Deploy edge computing gateways to receive data from smart meter terminals and environmental perception terminals and perform data alignment and preliminary processing.

[0054] The smart meter terminal collects the power consumption of the motor and air-conditioning unit once an hour and uploads the data to the edge computing gateway.

[0055] The environmental perception terminal collects environmental parameters in real time and uploads the data to the edge computing gateway.

[0056] After receiving the data, the edge computing gateway performs data alignment to ensure the temporal consistency of power data and environmental data. The gateway can then perform preliminary analysis on the data, such as calculating hourly energy consumption and identifying changing trends in environmental parameters.

[0057] Based on the analysis results of the edge computing gateway, the operating status of the motor and air conditioning unit can be adjusted in real time to achieve energy efficiency optimization. For example, when the ambient temperature is low, the set temperature of the air conditioning unit can be automatically lowered to reduce energy consumption.

[0058] At the same time, the analysis results can be uploaded to the cloud or central server for managers to conduct remote monitoring and decision support.

[0059] Step 120: Mark the regulation potential coefficient corresponding to the hourly electricity consumption impact data within the preset time period, and then use the collection time, device ID, environmental impact factors and regulation potential system to train the first machine learning algorithm to obtain the trained first machine learning algorithm; use the collection time, device ID, and device time-series power consumption to train the second machine learning algorithm to obtain the trained second machine learning algorithm; collect the household time-series power consumption on the user side of the preset regulation range within the preset time period, and then use the collection time and household time-series power consumption to train the third machine learning algorithm to obtain the trained third machine learning algorithm.

[0060] It should be noted that by marking the regulation potential coefficient corresponding to the hourly electricity consumption impact data within the preset time period, the regulation capacity of each device at different time points can be quantified. This helps to more accurately evaluate the regulation potential of the device and provide data support for subsequent regulation strategies.

[0061] By training the first machine learning algorithm with the acquisition time, device ID, environmental influencing factors and regulation potential coefficient, a model that can predict the device regulation potential can be obtained. This helps to formulate personalized regulation strategies based on the specific conditions of the device and environmental factors, and improve the accuracy and efficiency of regulation.

[0062] By training the second machine learning algorithm, the future power consumption of the device can be predicted using the collection time, device ID and device time series power consumption. This helps to plan energy distribution in advance, optimize power consumption plans and reduce energy consumption costs.

[0063] By collecting the household time-series electricity consumption of the user side in the preset control interval within the preset time period, and using the collection time and household time-series electricity consumption to train the third machine learning algorithm, a model of the user's electricity consumption behavior can be obtained. This helps to analyze the user's electricity consumption habits and needs and provide a basis for providing personalized energy services.

[0064] By combining the results of the three machine learning algorithms, comprehensive regulation of user-side adjustable resource equipment can be achieved. This includes adjusting the operating state of the equipment according to its regulation potential, optimizing energy distribution based on power consumption forecasts, and providing personalized energy recommendations based on user power consumption behavior. These measures work together to improve energy utilization efficiency and reduce energy consumption costs.

[0065] Specific examples: Assume that there is a smart park, which contains multiple user-side adjustable resource devices, such as smart air conditioners, smart lighting, etc. In order to achieve precise control and energy efficiency optimization of these devices, the park adopts the above method.

[0066] The park first collects hourly electricity consumption impact data within a preset time period (such as a week), including equipment time-series power consumption, environmental impact factors, etc.

[0067] Then, based on the performance parameters and historical data of the equipment, the regulation potential coefficient corresponding to the hourly electricity consumption impact data is marked.

[0068] Next, the first machine learning algorithm is trained using the collection time, device ID, environmental influencing factors, and regulation potential coefficient to obtain a model that can predict the regulation potential of the device.

[0069] The park continues to use the collection time, device ID, and device time-series power consumption to train the second machine learning algorithm to obtain a model that can predict the future power consumption of the equipment.

[0070] Through this model, the park can plan energy distribution in advance to ensure sufficient energy supply during peak electricity consumption periods.

[0071] The park collects the household time-series electricity consumption on the user side of the preset control range within the preset time period, and uses the collection time and household time-series electricity consumption to train the third machine learning algorithm.

[0072] Through this model, the park can analyze users' electricity usage habits and needs, providing a basis for providing personalized energy services.

[0073] Step 130: obtain the peak value of power consumption regulation within the future analysis interval, and obtain the predicted total power consumption of the preset regulation interval within the future analysis interval through the trained third machine learning algorithm; when the predicted total power consumption is greater than or equal to the peak value of power consumption regulation, obtain the specific actual value of the environmental influencing factor, and obtain the specific adjustment value of the adjustment potential coefficient through the preset adjustment rule, input the trained first machine learning algorithm, and obtain the device ID that meets the specific actual value and the specific adjustment value; input the device ID that meets the specific actual value and the specific adjustment value into the trained second machine learning algorithm to obtain the predicted time series power consumption corresponding to the device ID in the future analysis interval.

[0074] It should be noted that by obtaining the peak power consumption regulation and forecasting the total power consumption in the future analysis interval, the system can predict the peak and trough of power demand in advance, so as to carry out accurate power regulation. This helps to avoid power shortage or surplus and improve the stability and efficiency of the power system.

[0075] When the predicted total power consumption is greater than or equal to the power consumption regulation peak, the system obtains the specific actual value of the environmental influencing factor and obtains the specific adjustment value of the regulation potential coefficient according to the preset adjustment rules. This enables the system to make dynamic adjustments based on real-time environmental conditions, improving the flexibility and adaptability of regulation.

[0076] By inputting the device ID that meets the specific actual value and specific adjustment value into the trained first machine learning algorithm, the system can accurately identify the device that needs to be adjusted. This helps to achieve precise control at the device level and improve the pertinence and effectiveness of regulation.

[0077] By inputting the ID of the device that meets the conditions into the trained second machine learning algorithm, the system can predict the time-series power consumption of these devices in the future analysis interval. This helps to plan energy distribution in advance, optimize power consumption plans, and reduce energy consumption costs.

[0078] Specific examples: Assume that a city power grid system uses the above method to regulate power consumption in order to cope with the peak power consumption in summer.

[0079] The power grid system first obtains the peak value of power consumption regulation within the next 24 hours (analysis interval), which is predicted based on historical data and weather forecasts.

[0080] At the same time, through the trained third machine learning algorithm, the system predicts the total electricity consumption in the preset control area (such as commercial areas, residential areas, etc.) in the next 24 hours.

[0081] Assuming that the predicted total power consumption is greater than or equal to the power consumption control peak, the system obtains the specific actual values ​​of the current environmental influencing factors, such as temperature, humidity, wind speed, etc.

[0082] According to preset adjustment rules (such as increasing the adjustment potential coefficient of the air-conditioning equipment when the temperature is higher than a certain threshold), the system obtains the specific adjustment value of the adjustment potential coefficient.

[0083] The system inputs the device ID that meets the specific actual value and the specific adjustment value into the trained first machine learning algorithm to identify the equipment that needs to be adjusted, such as the air conditioning units in some commercial buildings.

[0084] The identified device IDs are input into the trained second machine learning algorithm, and the system predicts the sequential power consumption of these devices in the next 24 hours.

[0085] Based on these prediction results, the power grid system can plan energy distribution in advance, such as adjusting power generation plans and optimizing power grid scheduling, to ensure the stability and efficiency of power supply.

[0086] After obtaining the peak value of power consumption regulation in the future analysis interval and obtaining the predicted total power consumption of the preset regulation interval in the future analysis interval through the trained third machine learning algorithm, the method further includes: When the predicted total power consumption is less than or equal to the power consumption control peak value, the default program is maintained.

[0087] Preset adjustment rules, including: The specific adjustment value of the potential coefficient is adjusted in a manner of successively decreasing the preset value; wherein the preset adjustment rule includes a preset initial adjustment value.

[0088] It should be noted that when the predicted total power consumption is less than or equal to the power consumption regulation peak, the default program is maintained, which helps to maintain the stability of the system, avoid unnecessary regulation operations, and thus improve the operating efficiency of the system.

[0089] The preset adjustment rules, especially the specific adjustment value of the potential coefficient in a way of gradually decreasing the preset value, enable the system to dynamically adjust the operating status of the equipment according to the real-time power consumption and achieve precise control. This helps to optimize resource allocation and avoid energy waste.

[0090] The preset adjustment rules include preset initial adjustment values, which provide the system with a certain degree of flexibility and adaptability. The system can adjust the initial adjustment values ​​according to actual conditions to cope with different power demand and scenarios.

[0091] Through precise control and avoiding unnecessary control operations, the system can reduce control costs, including labor costs, equipment loss costs, etc.

[0092] Stable system operation and precise control strategies help improve user experience, such as maintaining a constant indoor temperature and avoiding frequent changes in lighting brightness.

[0093] Step 140: When the sum of all predicted time-series power consumption is greater than the difference between the predicted total power consumption and the power consumption regulation peak, a regulation instruction is sent to the user side corresponding to the device ID to obtain the device ID that returns the executable instruction; when the sum of power consumption corresponding to the device ID that returns the executable instruction is less than the difference between the predicted total power consumption and the power consumption regulation peak or a further downward adjustment instruction is received, the specific adjustment value of the adjustment potential coefficient is adjusted based on the preset adjustment rules.

[0094] It can be understood by those skilled in the art that when the sum of all predicted time-series power consumption is greater than the difference between the predicted total power consumption and the power consumption control peak, the system can promptly detect the potential risk of power overload and reduce power consumption by issuing control instructions to the user side corresponding to the device ID, thereby avoiding the occurrence of power overload.

[0095] By obtaining the device ID that returns executable instructions, it is possible to accurately identify which devices can participate in regulation, thereby achieving precise regulation. This helps optimize resource allocation and ensure the stability and efficiency of power supply.

[0096] When the summed power consumption corresponding to the device ID that returns the executable instruction is less than the difference between the predicted total power consumption and the power consumption regulation peak value or a further downward adjustment instruction is received, the specific adjustment value of the adjustment potential coefficient can be adjusted based on the preset adjustment rules.

[0097] By sending control instructions to the user side and obtaining the device ID that returns the executable instructions, the system can encourage users to participate in power control and improve user participation and satisfaction. At the same time, the system can also adjust the control strategy in time according to user feedback to better meet user needs.

[0098] In some embodiments, sending a control instruction to the user side corresponding to the device ID to obtain the device ID that returns the executable instruction specifically includes: Send control instructions to the user side corresponding to the device ID, and obtain execution instructions returned by the user side corresponding to the device ID; wherein the execution instructions are divided into executable and non-executable; Determines the device ID for which the executable command is returned.

[0099] The summed power consumption corresponding to the device ID that returns the executable command is less than the difference between the predicted total power consumption and the power consumption control peak value, or a further downward adjustment command is received, including: Get the predicted time series power consumption corresponding to the device ID that returns the executable instruction; Accumulate the predicted time series power consumption to obtain the summed power consumption; Calculate whether the total power consumption is less than the difference between the predicted total power consumption and the power consumption control peak value; Obtain further downward adjustment instructions through the preset information acquisition interface.

[0100] In addition, this application Figure 2 A user-adjustable resource control method system based on machine learning is provided in the embodiment of the present application. Figure 2 As shown, the system provided in the embodiment of the present application mainly includes: The collection module 210 is used to collect the power consumption impact data of the user-side adjustable resource equipment every hour within a preset time period; wherein the power consumption impact data includes: the equipment time series power consumption and environmental impact factors corresponding to the adjustable resource equipment, and the adjustable resource equipment includes a unique equipment ID.

[0101] The training module 220 is used to mark the regulation potential coefficient corresponding to the hourly electricity consumption impact data within a preset time period, and then use the collection time, device ID, environmental impact factors and regulation potential system to train the first machine learning algorithm to obtain a trained first machine learning algorithm; use the collection time, device ID, and device time-series power consumption to train the second machine learning algorithm to obtain a trained second machine learning algorithm; collect household time-series power consumption on the user side of a preset regulation range within a preset time period, and then use the collection time and household time-series power consumption to train the third machine learning algorithm to obtain a trained third machine learning algorithm.

[0102] The acquisition module 230 is used to obtain the peak power consumption regulation value within the future analysis interval, and to obtain the predicted total power consumption of the preset regulation interval within the future analysis interval through the trained third machine learning algorithm; when the predicted total power consumption is greater than or equal to the peak power consumption regulation value, the specific actual value of the environmental influencing factor is obtained, and the specific adjustment value of the regulation potential coefficient is obtained through the preset adjustment rule, and the trained first machine learning algorithm is input to obtain the device ID that meets the specific actual value and the specific adjustment value; the device ID that meets the specific actual value and the specific adjustment value is input into the trained second machine learning algorithm to obtain the predicted time series power consumption corresponding to the device ID in the future analysis interval; when the sum of all predicted time series power consumption is greater than the difference between the predicted total power consumption and the power regulation peak, a control instruction is issued to the user side corresponding to the device ID to obtain the device ID that returns an executable instruction.

[0103] The adjustment module 240 is used to adjust the specific adjustment value of the adjustment potential coefficient based on the preset adjustment rules when the summed power consumption corresponding to the device ID that returns the executable instruction is less than the difference between the predicted total power consumption and the power consumption regulation peak or a further downward adjustment instruction is received.

[0104] The adjustment module 240 includes an adjustment unit, which is used to obtain the predicted time series power consumption corresponding to the device ID that returns the executable instruction; accumulate the predicted time series power consumption to obtain the summed power consumption; calculate whether the summed power consumption is less than the difference between the predicted total power consumption and the power consumption control peak; and obtain further downward adjustment instructions through a preset information acquisition interface.

[0105] In addition, an embodiment of the present application further provides a non-volatile computer storage medium on which executable instructions are stored. When the executable instructions are executed, a user-adjustable resource control method based on machine learning as described above is implemented.

[0106] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A user-adjustable resource control method based on machine learning, characterized in that: The method comprises: Collect the power consumption impact data of the user-side adjustable resource equipment every hour within the preset time period; the power consumption impact data includes: the equipment time series power consumption and environmental impact factors corresponding to the adjustable resource equipment, and the adjustable resource equipment contains a unique equipment ID; Annotate the regulation potential coefficient corresponding to the hourly electricity consumption impact data within a preset time period, and then train the first machine learning algorithm using the collection time, device ID, environmental impact factors, and regulation potential system to obtain a trained first machine learning algorithm; Using the collection time, the device ID, and the device sequential power consumption, the second machine learning algorithm is trained to obtain a trained second machine learning algorithm; Collecting household time-series power consumption at the user side of a preset control interval within a preset time period, and then using the collection time and household time-series power consumption to train a third machine learning algorithm to obtain a trained third machine learning algorithm; Obtaining the peak value of power consumption regulation in the future analysis interval, and obtaining the predicted total power consumption of the preset regulation interval in the future analysis interval through the trained third machine learning algorithm; When the predicted total power consumption is greater than or equal to the power consumption regulation peak value, the specific actual value of the environmental influencing factor is obtained, and the specific adjustment value of the regulation potential coefficient is obtained through the preset adjustment rule, and the trained first machine learning algorithm is input to obtain the device ID that meets the specific actual value and the specific adjustment value; The device ID that meets the specific actual value and the specific adjustment value is input into the trained second machine learning algorithm to obtain the predicted time series power consumption corresponding to the device ID in the future analysis interval; When the sum of all predicted time-series power consumption is greater than the difference between the predicted total power consumption and the power consumption control peak value, a control instruction is sent to the user side corresponding to the device ID to obtain the device ID that returns the executable instruction; When the summed power consumption corresponding to the device ID that returns the executable instruction is less than the difference between the predicted total power consumption and the power consumption regulation peak value or a further downward adjustment instruction is received, the specific adjustment value of the adjustment potential coefficient is adjusted based on the preset adjustment rules.

2. The user-adjustable resource control method based on machine learning according to claim 1, characterized in that: Before collecting the power consumption impact data of the user-side adjustable resource equipment every hour within the preset time period, the method further includes: Obtain the original adjustable resource equipment information uploaded by the user through the preset data collection interface; The original adjustable resource equipment information includes at least: specific model information and rated power of the adjustable resource equipment; Determine whether the specific model information and rated power meet the preset limiting rules corresponding to the adjustable resource equipment; When the preset limiting rule is met, determining that the device corresponding to the original adjustable resource device information is an adjustable resource device; A unique device ID is randomly generated for the adjustable resource device and stored.

3. The user-adjustable resource control method based on machine learning according to claim 1, characterized in that: After obtaining the peak value of power consumption regulation in the future analysis interval and obtaining the predicted total power consumption of the preset regulation interval in the future analysis interval by using the trained third machine learning algorithm, the method further includes: When the predicted total power consumption is less than or equal to the power consumption control peak value, the default program is maintained.

4. The user-adjustable resource control method based on machine learning according to claim 1, characterized in that: Before collecting the power consumption impact data of the user-side adjustable resource equipment every hour within the preset time period, the method further includes: Configure smart meter terminals, environment sensing terminals and edge computing gateways for adjustable resource devices; Then, the smart meter terminal is used to collect the equipment's time-series power consumption and upload it to the edge computing gateway; Use environmental sensing terminals to collect environmental impact factors and upload them to the edge computing gateway; Through the edge computing gateway, data alignment is performed on the device’s sequential power consumption and environmental influencing factors.

5. The user-adjustable resource control method based on machine learning according to claim 1, characterized in that: Send control instructions to the user side corresponding to the device ID to obtain the device ID that returns the executable instructions, including: Send control instructions to the user side corresponding to the device ID, and obtain execution instructions returned by the user side corresponding to the device ID; wherein the execution instructions are divided into executable and non-executable; Determines the device ID for which the executable command is returned.

6. The user-adjustable resource control method based on machine learning according to claim 1, characterized in that: Preset adjustment rules, including: The specific adjustment value of the potential coefficient is adjusted in a manner of successively decreasing the preset value; wherein the preset adjustment rule includes a preset initial adjustment value.

7. The user-adjustable resource control method based on machine learning according to claim 1, characterized in that: The summed power consumption corresponding to the device ID that returns the executable command is less than the difference between the predicted total power consumption and the power consumption control peak value, or a further downward adjustment command is received, including: Get the predicted time series power consumption corresponding to the device ID that returns the executable instruction; Accumulate the predicted time series power consumption to obtain the summed power consumption; Calculate whether the total power consumption is less than the difference between the predicted total power consumption and the power consumption control peak value; Obtain further downward adjustment instructions through the preset information acquisition interface.

8. A user-adjustable resource control method system based on machine learning, characterized in that: The system comprises: A collection module is used to collect the power consumption impact data of the user-side adjustable resource equipment every hour within a preset time period; wherein the power consumption impact data includes: the equipment time sequence power consumption and environmental impact factors corresponding to the adjustable resource equipment, and the adjustable resource equipment includes a unique equipment ID; A training module is used to mark the adjustment potential coefficient corresponding to the hourly electricity consumption impact data within a preset time period, and then use the collection time, device ID, environmental impact factors and adjustment potential system to train the first machine learning algorithm to obtain a trained first machine learning algorithm; use the collection time, device ID, and device time-series power consumption to train the second machine learning algorithm to obtain a trained second machine learning algorithm; collect the household time-series power consumption on the user side of the preset regulation interval within the preset time period, and then use the collection time and household time-series power consumption to train the third machine learning algorithm to obtain a trained third machine learning algorithm; An acquisition module is used to obtain a peak value of power consumption regulation within a future analysis interval, and to obtain a predicted total power consumption of a preset regulation interval within the future analysis interval through a trained third machine learning algorithm; when the predicted total power consumption is greater than or equal to the peak value of power consumption regulation, the specific actual value of the environmental influencing factor is obtained, and the specific adjustment value of the regulation potential coefficient is obtained through a preset adjustment rule, and the trained first machine learning algorithm is input to obtain a device ID that satisfies the specific actual value and the specific adjustment value; the device ID that satisfies the specific actual value and the specific adjustment value is input into a trained second machine learning algorithm to obtain the predicted time series power consumption corresponding to the device ID in the future analysis interval; when the sum of all predicted time series power consumption is greater than the difference between the predicted total power consumption and the power consumption regulation peak, a control instruction is issued to the user side corresponding to the device ID to obtain the device ID that returns an executable instruction; The adjustment module is used to adjust the specific adjustment value of the adjustment potential coefficient based on the preset adjustment rules when the summed power consumption corresponding to the device ID that returns the executable instruction is less than the difference between the predicted total power consumption and the power consumption regulation peak or when a further downward adjustment instruction is received.

9. The user-adjustable resource control method system based on machine learning according to claim 8, characterized in that: The regulating module comprises a regulating unit, Used to obtain the predicted time series power consumption corresponding to the device ID that returns the executable instruction; Accumulate the predicted time series power consumption to obtain the summed power consumption; Calculate whether the total power consumption is less than the difference between the predicted total power consumption and the power consumption control peak value; Obtain further downward adjustment instructions through the preset information acquisition interface.

10. A non-volatile computer storage medium, characterized in that: Computer instructions are stored thereon, and when the computer instructions are executed, they implement a user-adjustable resource control method based on machine learning as described in any one of claims 1-7.

Images