Building load pressure drop collaborative optimization allocation method and system for virtual power plant

Abstract

The application discloses a building load pressure drop cooperative optimization distribution method and system for a virtual power plant, and relates to the technical field of load pressure drop distribution. The method comprises the following steps: S1, obtaining a scheduling time step, collecting multi-dimensional regulation and control characteristic data in real time, constructing a comfort evaluation model, and performing data preprocessing; S2, evaluating the regulation priority of equipment that can participate in load pressure drop regulation, and generating a regulation priority list; S3, performing pressure drop regulation, and performing pressure drop regulation decomposition and feedback among multiple power utilization equipment; S4, judging overall cooperative regulation performance, identifying abnormal indexes, and performing optimization; and S5, monitoring multi-dimensional regulation and control characteristic data, regulation priority, pressure drop regulation decomposition results and overall cooperative regulation performance, improving a pressure drop distribution strategy, and performing algorithm optimization. The method solves the problem that, in the prior art, when a pressure drop strategy is dynamically adjusted under multi-target constraints, the response capability of virtual power plant regulation and control and user experience are difficult to be considered, and safe and efficient pressure drop distribution is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of load pressure drop allocation, in particular to a building load pressure drop collaborative optimization allocation method and system for a virtual power plant. BACKGROUND

[0002] With the continuous growth of urban electricity load and the increase of renewable energy proportion, the power system has increasingly high requirements for flexible load regulation and demand response. Load pressure drop, as an important means of virtual power plant and intelligent building energy efficiency management, can quickly and safely realize power reduction and electricity behavior optimization of building energy units under peak load, tight supply and demand, or grid regulation demand.

[0003] For example, the invention patent with publication number CN117575206A discloses a power grid load pressure drop capacity rapid evaluation method and system, and specifically relates to the technical field of power grid load evaluation, which comprises the following steps: step one, obtaining user profile information and user load information; step two, developing a load pressure drop capacity evaluation module; step three, classifying according to industry information in the user marketing profile; step four, evaluating the load pressure drop capacity of users in each industry; and step five, dynamically updating the load pressure drop capacity of the whole society according to the classification evaluation results. The present application relies on a new type of power load management system, collects data information from marketing business systems, power utilization information collection systems, and dispatching control systems, and can quickly evaluate the load pressure drop capacity of different types of power users by classifying the load resources. The present application can provide guidance for demand-side power supply during a severe power supply and demand situation, and can also provide a basis for calculating the load adjustment potential of each industry, thereby supporting the normalization of load management.

[0004] However, in the process of implementing the technical scheme of the present application, the present application finds that the above-mentioned technology at least has the following technical problems:

[0005] Building load pressure drop is usually realized by reducing the power of air conditioners, lighting and non-critical electrical equipment. However, excessive pressure drop may cause a decrease in user comfort and damage to equipment life. The existing technology is difficult to dynamically adjust the pressure drop strategy under multi-objective constraints, fully exert the rapid response capability of building load to the regulation and control demand of virtual power plant, and at the same time, take into account user experience and equipment operation health, to realize safe, reasonable and efficient pressure drop allocation.

[0006] Therefore, in view of the above problems, there is an urgent need for a building load pressure drop collaborative optimization allocation method and system for a virtual power plant. SUMMARY

[0007] Technical problems solved

[0008] In view of the deficiencies of the prior art, the building load pressure drop collaborative optimization and distribution method and system for a virtual power plant are provided, and the problem that the prior art is difficult to balance the response capability of virtual power plant regulation and control and user experience when dynamically adjusting the pressure drop strategy under multi-objective constraints, and to realize safe and efficient pressure drop distribution is solved.

[0009] Technical scheme

[0010] To achieve the above object, the building load pressure drop collaborative optimization and distribution method for a virtual power plant is implemented by the following technical scheme: S1, a scheduling time step is obtained, multi-dimensional regulation and control feature data are collected in real time, a comfort evaluation model is constructed, and the multi-dimensional regulation and control feature data are preprocessed; S2, according to the multi-dimensional regulation and control feature data, equipment that can participate in load pressure drop regulation is identified, and in combination with the preprocessed multi-dimensional regulation and control feature data and the comfort evaluation model, the regulation and control priority of each equipment is evaluated, and a regulation priority list is generated according to the regulation and control priority; S3, the regulation priority list is received for pressure drop regulation, and in combination with the preprocessed multi-dimensional regulation and control feature data and the regulation and control priority, pressure drop regulation decomposition among multiple power equipment is performed, and the pressure drop regulation decomposition result is fed back; S4, according to the feedback of the pressure drop regulation decomposition result, the overall collaborative regulation and control performance is judged by comprehensively combining the pressure drop regulation decomposition result and the multi-dimensional regulation and control feature data before and after pressure drop regulation, abnormal indexes are identified according to the overall collaborative regulation and control performance, and optimization is performed; S5, the multi-dimensional regulation and control feature data, the regulation and control priority, the pressure drop regulation decomposition result and the overall collaborative regulation and control performance are continuously monitored, the pressure drop distribution strategy is improved, and algorithm tuning is performed.

[0011] Further, the specific process of acquiring the scheduling time step, collecting multi-dimensional regulation feature data in real time, constructing a comfort evaluation model, and data preprocessing of the multi-dimensional regulation feature data is as follows: by deploying multiple types of sensors at key nodes of the building, real-time collection of power load data, electrical parameters, and environmental parameters is performed, wherein the electrical parameters include harmonic content, voltage value, and current value, and the environmental parameters include temperature, humidity, lighting brightness, and air quality of each area; the total target value of the load pressure drop required to be adjusted issued by the virtual power plant is recorded in real time; at the same time, the scheduling time step is acquired, which is defined as the time interval of each round of load pressure drop regulation, distribution, and execution, serving as the time reference of each process; the standard load regulation test instruction is issued to the equipment, the real-time running data of the equipment is collected synchronously, the response time from the instruction issuance to the actual load change reaching the target and the load change rate are recorded, and the response speed of the equipment is calculated, while the controlled power regulation process of the equipment is analyzed, the maximum rate of actual power change per unit time is calculated and selected as the climbing ability of the equipment; the rated minimum power constant of each equipment is acquired, and the current running power is collected in real time by the on-site sensor, and the difference between the current running power and the rated minimum power constant is used to calculate the remaining adjustable space; based on the multi-dimensional environmental parameters collected in real time, a multi-parameter environmental state data set is constructed, principal component analysis algorithm is adopted to cluster and reduce the dimension of the multi-parameter environmental state data set, and the natural distribution interval of the environmental parameters is divided, the environmental state represented by each cluster center is taken as a typical comfort interval, the principal component score value of each environmental parameter and the cluster center is taken as the environmental comfort reference value and is normalized to construct a comfort evaluation model, and the comfort scores corresponding to different environmental parameters are output; the total target value of the load pressure drop, the scheduling time step, the power load data, the electrical parameters, the environmental parameters, the response time, the response speed, the climbing ability, the running power, the remaining adjustable space, and the comfort scores are recorded as multi-dimensional regulation feature data; the multi-dimensional regulation feature data is time-aligned, standardized, and normalized, and the median filtering method is used to suppress noise, abnormal values are removed through physical range verification, Hampel filtering, and sliding window standard score detection, and Kalman filtering algorithm is used to interpolate missing data; and a load pressure drop coordination database is constructed to store the multi-dimensional regulation feature data.

[0012] Further, according to the multi-dimensional regulation characteristic data, the device that can participate in the load pressure drop regulation is identified, and the specific process of evaluating the regulation priority of each device is combined with the pre-processed multi-dimensional regulation characteristic data and the comfort evaluation model: obtaining the multi-dimensional regulation characteristic data, identifying all devices that can participate in the load pressure drop regulation at present; for each device, calculating the response capability value by dividing the climbing ability by the sum of the response time and the constant one; calculating the adjustable space proportion by dividing the remaining down-regulation space by the current operating power; obtaining the harmonic content and voltage value in the historical period, respectively calculating the harmonic content mean value and the harmonic content standard deviation, and the voltage value mean value and the voltage value standard deviation, performing standard score normalization, and calculating the sum of the normalized values of the harmonic content and the voltage value as the electrical disturbance risk value; multiplying the response capability value, the adjustable space proportion and the current comfort score, and dividing by the sum of the electrical disturbance risk value and the constant one to obtain the device regulation priority.

[0013] Further, the specific process of generating a regulation priority list according to the regulation priority is: arranging all devices that can participate in the load pressure drop regulation in descending order according to the device regulation priority to generate a regulation priority list; at the same time, combining the total load pressure drop target value, counting the remaining down-regulation space of all devices that can participate in the load pressure drop regulation, and judging whether the overall achievable pressure drop capacity meets the target demand; according to the priority list and the remaining down-regulation space and the climbing ability of each device, the pressure drop instruction is sequentially assigned to each device; if the remaining down-regulation space of the device with the highest device regulation priority is insufficient to cover the total load pressure drop target value, the pressure drop instruction is assigned to the next level device, until the overall pressure drop target is achieved or all devices run out of space, then the assignment is stopped; the device regulation priority is written into the load pressure drop coordination database.

[0014] Further, the specific process of receiving the regulation priority list for pressure drop regulation, and combining the pre-processed multi-dimensional regulation characteristic data and the regulation priority to decompose the pressure drop regulation among multiple power consumption devices is: receiving the current regulation priority list, and according to the real-time physical constraint and the total load pressure drop target value, the load pressure drop instruction is issued to each device; for each device, obtaining the multi-dimensional regulation characteristic data and the device regulation priority; multiplying the climbing ability by the scheduling time step to obtain the dynamic limit allocation value; at the same time, summing up the device regulation priorities of all devices to obtain the priority sum, calculating the ratio of the device regulation priority of device k to the priority sum, and multiplying the ratio by the total load pressure drop target value to obtain the priority weighted allocation value; comparing the remaining down-regulation space, the dynamic limit allocation value and the priority weighted allocation value of device k, and taking the minimum value among the three as the load pressure drop allocation value.

[0015] Further, the specific process of feeding back the pressure drop regulation decomposition result is: according to the calculated load pressure drop distribution value of each device, generating specific regulation and control instructions for each device in this round, and issuing them to each target device in real time through the controller; collecting and recording the state information of each device after regulation in real time, including: actual power change, device operating state, environmental parameters, electrical parameters, and whether the device responds to the instruction; comparing the actual power change of the device with the load pressure drop distribution value to determine whether the device completes the load pressure drop according to the plan, and if the device response is found to be abnormal, it is marked and fed back; recording the load pressure drop distribution value and execution log in each round of regulation process, including the instruction issuing time, response feedback time and actual regulation effect, and uploading them to the load pressure drop coordination database.

[0016] Further, according to the feedback of the pressure drop regulation decomposition result, the specific process of comprehensively judging the overall coordinated control performance based on the pressure drop regulation decomposition result and the multi-dimensional regulation and control characteristic data before and after the pressure drop regulation is: receiving the load pressure drop distribution value and the actual regulation effect of each device, analyzing the deviation of the response effect of each device from the target in real time, and determining the overall load drop execution achievement; based on the total number of all devices participating in the load pressure drop regulation, calculating the load pressure drop distribution value of all devices and summing them up to obtain the load pressure drop total value, subtracting the constant one from the ratio of the load pressure drop total value to the load pressure drop target value, and squaring to obtain the load pressure drop target implementation error; obtaining the electrical parameters of all devices before and after the load pressure drop regulation and the comfort score before and after the load pressure drop regulation from the load pressure drop coordination database; calculating the difference between the voltage value before the load pressure drop regulation and the voltage value after the load pressure drop regulation and taking the absolute value to obtain the voltage change, dividing the voltage change by the rated working voltage constant and squaring to obtain the voltage relative fluctuation value, summing up the voltage relative fluctuation values of all devices to obtain the voltage fluctuation total value; at the same time, calculating the standard deviation of the comfort score of the region where the device is located in the historical period, calculating the difference between the comfort score before the load pressure drop regulation and the comfort score after the load pressure drop regulation and taking the absolute value to obtain the comfort change, dividing the comfort change by the comfort score standard deviation and squaring to obtain the comfort fluctuation value, summing up the comfort fluctuation values of all devices to obtain the comfort fluctuation total value; adding the load pressure drop target implementation error, the voltage fluctuation total value and the comfort fluctuation total value to obtain the multi-target coordinated control performance evaluation value.

[0017] Further, the specific process of identifying abnormal indicators and optimizing according to the overall coordinated regulation performance is: after each load pressure drop regulation period ends, the multi-objective coordinated regulation performance evaluation value and three indicators are calculated, the three indicators include: pressure drop target implementation error, voltage fluctuation total, and comfort fluctuation total, the change trend of the multi-objective coordinated regulation performance evaluation value and the three indicators is compared with history, and the abnormality in the pressure drop completion degree, electrical disturbance and comfort risk is identified; if the multi-objective coordinated regulation performance evaluation value is greater than the abnormal threshold and the three indicators are abnormal, a risk warning is issued, and the abnormal equipment, time period and indicator item are recorded; the multi-objective coordinated regulation performance evaluation value and the three indicators are written into the load pressure drop coordination database, machine learning and rule engine are used for regular analysis, the multi-objective coordinated regulation performance evaluation value algorithm is updated, and the multi-objective coordinated regulation performance evaluation value change trend report, abnormal indicator analysis, parameter adjustment record and optimization log are periodically generated and archived.

[0018] Further, the specific process of continuously monitoring multi-dimensional regulation feature data, regulation priority, pressure drop regulation decomposition result and overall coordinated regulation performance, improving pressure drop allocation strategy and algorithm tuning is: based on the real-time collected multi-dimensional regulation feature data, the environmental and electrical changes caused by the load pressure drop regulation are monitored; the history trend and abnormality of the multi-objective coordinated regulation performance evaluation value and the three indicators are analyzed regularly, the next round of load pressure drop regulation parameters are adjusted according to the found abnormality, if the comfort fluctuation total is continuously higher than the comfort fluctuation threshold, the regulation amplitude of the equipment in the related area is reduced in the next regulation; if the voltage fluctuation total is higher than the voltage fluctuation threshold, the equipment with a voltage relative fluctuation value greater than the safety fluctuation threshold is limited to participate in the pressure drop; at the same time, the feedback of user satisfaction and environmental parameters is fused, and the genetic algorithm is used for continuous optimization and self-learning iteration of the equipment regulation priority and load pressure drop allocation value algorithm; through the data-driven closed-loop feedback, the adaptive evolution of the load pressure drop regulation and allocation strategy is realized, and the coordinated optimization closed loop of the load pressure drop regulation is constructed.

[0019] The second aspect of the present application provides a building load pressure drop coordinated optimization and distribution system for a virtual power plant, comprising: a multi-source data acquisition and environment monitoring module, configured to obtain a scheduling time step, acquire multi-dimensional regulation and control characteristic data in real time, construct a comfort evaluation model, and perform data preprocessing on the multi-dimensional regulation and control characteristic data; an adjustable resource identification and priority evaluation module, configured to identify devices that can participate in load pressure drop regulation according to the multi-dimensional regulation and control characteristic data, evaluate the regulation priority of each device in combination with the preprocessed multi-dimensional regulation and control characteristic data and the comfort evaluation model, and generate a regulation priority list according to the regulation priority; a load pressure drop instruction decomposition and distribution module, configured to receive the regulation priority list for pressure drop regulation, and perform pressure drop regulation decomposition among multiple power utilization devices in combination with the preprocessed multi-dimensional regulation and control characteristic data and the regulation priority, and feed back the pressure drop regulation decomposition result; a multi-target dynamic regulation and safety constraint module, configured to identify abnormal indicators and perform optimization according to the feedback of the pressure drop regulation decomposition result, and judge the overall coordinated regulation performance by combining the pressure drop regulation decomposition result and the multi-dimensional regulation and control characteristic data before and after pressure drop regulation; and a user environment guarantee and optimization module, configured to continuously monitor the multi-dimensional regulation and control characteristic data, the regulation priority, the pressure drop regulation decomposition result and the overall coordinated regulation performance, improve the pressure drop distribution strategy and perform algorithm tuning.

[0020] Advantages

[0021] The present application has the following advantages:

[0022] (1) The present application realizes comprehensive digital expression of environment perception, energy utilization state and device capacity by acquiring power utilization load, electrical parameters and environmental parameters in real time, and using unsupervised machine learning algorithm to cluster and reduce the dimension of multi-parameter environmental state, and constructs a comfort evaluation model, thereby providing accurate and dynamic data basis for pressure drop regulation.

[0023] (2) The present application objectively evaluates the priority of all adjustable devices by using the multi-dimensional characteristics of climbing ability, response speed, remaining adjustable space, electrical disturbance risk and comfort score obtained by real-time acquisition and historical analysis, dynamically generates a regulation priority list, supports fine decomposition and distribution of pressure drop instructions, avoids the problems of extensive and lagging response of traditional pressure drop regulation resource allocation, and improves the scientificity and flexibility of overall pressure drop response.

[0024] (3) The present application constructs a multi-target coordinated regulation performance evaluation value based on pressure drop target implementation error, voltage fluctuation sum and comfort fluctuation sum in the process of load pressure drop distribution, forms a dynamic closed-loop monitoring and abnormal self-checking mechanism throughout the whole process, adjusts the regulation strategy and distribution parameters according to the real-time evaluation result, effectively prevents and controls the electrical instability and user comfort damage caused by excessive pressure drop, and balances the regulation safety and user experience.

[0025] (4) Based on multi-target cooperative control performance evaluation value, environmental feedback and user satisfaction feedback, the genetic algorithm is adopted to continuously adaptively adjust the equipment control priority and pressure drop distribution value, realize the dynamic evolution of the pressure drop regulation parameter and algorithm, improve the adaptability and robustness of the pressure drop regulation, and promote the continuous optimization of the performance and the continuous improvement of the intelligent level.

[0026] Of course, implementing any product of the present application does not necessarily require all the advantages described above to be achieved simultaneously. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 A flow chart of a building load pressure drop cooperative optimization distribution method for a virtual power plant;

[0028] Figure 2 A system structure diagram of a building load pressure drop cooperative optimization distribution system for a virtual power plant;

[0029] Figure 3 A radar chart of equipment control priority for each control feature;

[0030] Figure 4 A column chart of equipment control priority. DETAILED DESCRIPTION

[0031] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. As understood by those skilled in the art, the described embodiments are only some of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0032] Please refer to Figures 1-4 The embodiments of the present application provide a technical solution: a building load pressure drop cooperative optimization distribution method and system for a virtual power plant, such as Figure 1As shown, the method comprises the following steps: S1, obtaining a scheduling time step, collecting multi-dimensional regulation and control characteristic data in real time, constructing a comfort evaluation model, and performing data preprocessing on the multi-dimensional regulation and control characteristic data; S2, identifying devices that can participate in load pressure drop regulation according to the multi-dimensional regulation and control characteristic data, and combining the preprocessed multi-dimensional regulation and control characteristic data and the comfort evaluation model to evaluate the regulation priority of each device, and generating a regulation priority list according to the regulation priority; S3, receiving the regulation priority list for pressure drop regulation, and combining the preprocessed multi-dimensional regulation and control characteristic data and the regulation priority to perform pressure drop regulation decomposition among multiple power consumption devices, and feeding back the pressure drop regulation decomposition result; S4, according to the feedback of the pressure drop regulation decomposition result, comprehensively judging the overall collaborative regulation and control performance by combining the pressure drop regulation decomposition result and the multi-dimensional regulation and control characteristic data before and after the pressure drop regulation, identifying abnormal indicators according to the overall collaborative regulation and control performance and optimizing the abnormal indicators; S5, continuously monitoring the multi-dimensional regulation and control characteristic data, the regulation priority, the pressure drop regulation decomposition result and the overall collaborative regulation and control performance, improving the pressure drop allocation strategy and performing algorithm tuning.

[0033] In particular, the scheduling time step is obtained, the multi-dimensional regulation feature data is collected in real time, the comfort evaluation model is constructed, and the specific process of data preprocessing of the multi-dimensional regulation feature data is as follows: by deploying multiple types of sensors at key nodes of the building, the multiple types of sensors include electric energy meters, voltage and current transformers, environmental temperature and humidity sensors, air quality detectors, and illumination intensity sensors, which can cover key circuits and regional spaces, realize high-frequency and high-precision real-time collection of multi-source data such as power load data, electrical parameters and environmental parameters, wherein the electrical parameters include harmonic content, voltage value and current value, and the environmental parameters include temperature, humidity, lighting brightness and air quality of each region; the total target value of the load pressure drop required to be adjusted issued by the virtual power plant is recorded in real time; at the same time, the scheduling time step is obtained, which is defined as the time interval of each round of load pressure drop regulation, distribution and execution, serving as the time reference of each process, and being used to control the execution frequency and feedback update rate of the scheduling strategy; the standard load regulation test instruction is issued to the device, the real-time running data of the device is collected synchronously, the response time and load change rate from the instruction issuance to the actual load change reaching the target are recorded, and the response speed of the device is calculated, that is, by analyzing the whole process of load change, the dynamic regulation capability of the device after receiving the instruction is quantified, and the controlled power regulation process of the device is analyzed, the maximum rate of actual power change per unit time is calculated and selected as the climbing capability of the device, the climbing capability is the maximum safe load change rate that can be achieved per unit time, and reflects the dynamic limit of the device participating in the pressure drop regulation; the rated minimum power constant of each device is obtained, and the current running power is collected in real time by the on-site sensor, and the difference between the current running power and the rated minimum power constant is used to calculate the remaining down-regulation space, which represents the available load capacity that can be further down-regulated by the device under the condition of ensuring safe operation; based on the real-time collected multi-dimensional environmental parameters, a multi-parameter environmental state data set is constructed, a principal component analysis algorithm, i.e., a PCA algorithm, is adopted to reduce the dimension and extract the features of the high-dimensional environmental data, the effectiveness of the feature expression and the model calculation efficiency are improved, the multi-parameter environmental state data set is clustered and dimensionally analyzed, the natural distribution interval of the environmental parameters is divided, the environmental parameter space is divided into different state categories by the clustering algorithm, the environmental state represented by each clustering center is taken as a typical comfort interval, the principal component score values of each environmental parameter and the clustering center are taken as environmental comfort reference values and normalized, a comfort evaluation model is constructed, and the comfort scores corresponding to different environmental parameters are output; the total target value of the load pressure drop, the scheduling time step, the power load data, the electrical parameters, the environmental parameters, the response time, the response speed, the climbing capability, the running power, the remaining down-regulation space and the comfort scores are recorded as multi-dimensional regulation feature data; the multi-dimensional regulation feature data is time-aligned, standardized and normalized, and the median filtering method is adopted to suppress noise, effectively suppressing outliers and short-term abnormalities;The abnormal values are removed through physical range check, Hampel filtering and sliding window standard score detection, and Kalman filtering algorithm is used to fill in the missing data. When there is data missing, the data is smoothed and predicted to complete. A load pressure drop collaborative database is constructed to store multi-dimensional regulation and control characteristic data, support high-frequency, multi-dimensional, structured and traceable storage, and realize efficient support for subsequent regulation and control optimization and traceability analysis.

[0034] In the embodiment, through high-frequency multi-source sensor deployment and full-process data acquisition, automatic feature extraction, intelligent anomaly detection and normalization processing, accurate perception and digital expression of building load, electrical environment and comfort state are realized. Through principal component analysis and clustering algorithms, the scientificity and adaptability of environmental data modeling are effectively improved, and the objectivity and dynamic optimization ability of comfort evaluation are guaranteed. Robust denoising, anomaly removal and missing data filling mechanisms are introduced in the whole link, which significantly enhances the data reliability and stability of regulation and control decision of the system in complex working conditions. Finally, the efficient storage and management of multi-dimensional regulation and control characteristic data provide a solid data foundation and technical support for subsequent pressure drop strategy allocation, equipment priority optimization and global performance optimization.

[0035] Specifically, according to the multi-dimensional regulation characteristic data, the devices that can participate in the load pressure drop regulation are identified, and the specific process of evaluating the regulation priority of each device is combined with the pre-processed multi-dimensional regulation characteristic data and the comfort evaluation model: obtaining the multi-dimensional regulation characteristic data, identifying all devices that can participate in the current load pressure drop regulation, and selecting the target devices that can participate in the current regulation task by judging the current running state of the devices and the safe working range of the devices; for each device, the response capability value is obtained by dividing the climbing ability by the response time and the constant one, the climbing ability and the response time measure the fast response capability of the device to the regulation instruction, and the constant one is used to prevent the denominator from being zero and to standardize the response capability value; the adjustable space ratio is obtained by calculating the ratio of the remaining down-regulation space to the current running power, that is, the remaining down-regulation space is the effective load interval that can be continuously down-regulated by the device, and the adjustable space ratio reflects the actual adjustable flexibility of the device under the current state; the harmonic content and the voltage value in the historical period are obtained, the harmonic content average and the harmonic content standard deviation and the voltage value average and the voltage value standard deviation are calculated respectively, the z-score normalization is performed, the z-score normalization is calculated by each point and the historical average and the standard deviation, which is convenient for horizontal comparison of electrical indicators at different times and different devices, and the sum of the normalized values of the harmonic content and the voltage value is calculated as the electrical disturbance risk value, which comprehensively reflects the influence on the stability of the electrical system in the regulation process, and the higher the value, the greater the disturbance risk; the regulation priority of the device is obtained by multiplying the response capability value, the adjustable space ratio and the current comfort score by the sum of the electrical disturbance risk value and the constant one, which realizes the comprehensive evaluation of the multi-target capability of the device, ensures that the priority ranking considers the regulation response and the adjustable flexibility, and takes into account the environmental comfort and the electrical safety, thereby providing a scientific decision basis for subsequent pressure drop allocation.

[0036] wherein the specific formula of the device regulation priority is:

[0037] ;

[0038] In the formula, the device regulation priority is used to comprehensively evaluate the current regulation priority of each adjustable device; the greater the device regulation priority, the stronger the dynamic adjustable capability of the device, the faster the response, the larger the space, the higher the data quality, and the smaller the influence of the regulation process on the electrical and comfort risks, and the device should be selected preferentially in the pressure drop allocation; the climbing ability of the device k reflects the instantaneous dynamic regulation capability of the device, and the greater the climbing ability, the more suitable the device is for bearing larger and faster load regulation; the response time of the device k is shorter, the faster the response, the stronger the ability of the device to adapt to dynamic demand, and the higher the device regulation priority, and the constant one is added to prevent the denominator from being zero and to ensure the smoothness of the value; represents the remaining down-regulation space of the device k, the greater the remaining down-regulation space, the greater the adjustment potential; represents the current running power of the device k, preventing a virtual high priority of an absolute space but a large scale of the device itself; represents the comfort score of the area where the device k is currently located, the higher the comfort score, the more comfortable the environment, indicating that there is a larger adjustment margin; the lower the comfort score, the closer to the lower limit of comfort, the device adjustment brings an increased risk of discomfort, and the priority is reduced; represents the electrical disturbance risk value of the device k, the greater the electrical disturbance risk value, the greater the denominator, reducing the priority of the device regulation, and ensuring overall electrical safety.

[0039] In this embodiment, Table 1 is a device regulation priority data table, which records in detail the ramping ability, response time, current running power, remaining down-regulation space, comfort score, electrical disturbance risk value and device regulation priority of different devices. Among them, the ramping ability corresponding to the device 1 is 1.2, the response time is 8, the current running power is 5.0, the remaining down-regulation space is 2.0, the comfort score is 0.95, the electrical disturbance risk value is 0.12, and the device regulation priority is 0.045; the ramping ability corresponding to the device 2 is 0.8, the response time is 12, the current running power is 3.5, the remaining down-regulation space is 1.0, the comfort score is 0.90, the electrical disturbance risk value is 0.08, and the device regulation priority is 0.0146; the ramping ability corresponding to the device 3 is 1.6, the response time is 5, the current running power is 6.0, the remaining down-regulation space is 3.0, the comfort score is 0.97, the electrical disturbance risk value is 0.20, and the device regulation priority is 0.108; the ramping ability corresponding to the device 4 is 0.5, the response time is 15, the current running power is 2.8, the remaining down-regulation space is 0.6, the comfort score is 0.88, the electrical disturbance risk value is 0.18, and the device regulation priority is 0.005; the ramping ability corresponding to the device 5 is 1.0, the response time is 7, the current running power is 4.0, the remaining down-regulation space is 1.5, the comfort score is 0.92, the electrical disturbance risk value is 0.10, and the device regulation priority is 0.039.

[0040] Table 1 Device Regulation Priority Data Table

[0041]

[0042] As Figure 3As shown, this is a normalized radar chart of the control characteristics for each equipment control priority. It displays the normalized data distribution and differences of five devices across six key indicator dimensions: ramp-up capability, response time, current operating power, remaining adjustable range, comfort score, and electrical disturbance risk value. Each colored line represents one device; the further the value on each indicator axis is from the center, the larger the value of that indicator. (Based on Table 1 and...) Figure 3 It can be seen that Equipment 3 achieves high values ​​close to 1 in several key indicators across six dimensions, including ramp-up capability, remaining adjustable capacity, current operating power, comfort score, and electrical disturbance risk value. Its overall polygon size is significantly expanded, indicating excellent comprehensive performance. Equipment 4 has the highest value in response time but is at the minimum or low level in most other indicators. Its polygon shape is contracted, resembling a ray. Equipment 4's control priority is only 0.005, the lowest in the group, indicating the weakest control capability. Equipment 1 and Equipment 5 have relatively balanced distributions across most indicators, without obvious weaknesses or extremely high values. Their normalized radar graph polygons show moderate expansion, corresponding to moderate control priorities, making them suitable as alternative equipment for load drop regulation.

[0043] like Figure 4 The chart shown is a bar chart of equipment control priority. The horizontal axis represents the equipment number, and the vertical axis represents the corresponding equipment control priority. The height of each bar reflects the equipment's ranking in this round of priority evaluation; the higher the value, the stronger the equipment's overall control capability, and the more suitable it is to take priority in load voltage drop regulation tasks. (Based on Table 1 and...) Figure 4 It can be seen that Equipment 3 has the highest priority for load control, possessing comprehensive advantages across the six key indicator dimensions, making it the best choice for priority allocation of load reduction in this round. Equipment 1 and Equipment 5 have the second highest priority for load control, indicating that they also perform well in key indicators and can be considered as secondary choices for load reduction regulation. Equipment 2 and Equipment 4 have significantly lower priority for load control, with Equipment 4 having the lowest priority, implying weak control response capabilities or limited constraints, making them unsuitable as the main implementers of load reduction in this round.

[0044] This implementation plan achieves scientific screening and dynamic priority assessment of building electrical equipment participating in load voltage drop regulation through in-depth fusion and analysis of multi-dimensional control characteristic data. By comprehensively considering equipment responsiveness, adjustability flexibility, environmental comfort, and electrical safety risks, the prioritization process balances rapid response and efficient resource utilization while ensuring user experience. This improves the accuracy, rationality, and intelligence of voltage drop regulation, laying a solid data foundation and decision support for subsequent coordinated voltage drop command allocation and multi-objective control optimization.

[0045] Specifically, the specific process of generating the regulation priority list according to the regulation priority is as follows: all devices that can participate in load pressure drop regulation are arranged in descending order according to the device regulation priority, so that devices with high priority are given priority to participate in pressure drop regulation, and a regulation priority list is generated; at the same time, in combination with the total target value of load pressure drop, the remaining down-regulation space of all devices that can participate in load pressure drop regulation is counted, which helps to comprehensively grasp the response ability of this round of pressure drop, and to judge whether the ability of the overall pressure drop can meet the target demand; according to the priority list and the remaining down-regulation space and the climbing ability of each device, the pressure drop instructions are allocated to each device in turn, and the physical regulation limit of the device itself is satisfied during the allocation to ensure the safety and executability of the regulation instructions; if the remaining down-regulation space of the device with the highest regulation priority is insufficient to cover the total target value of load pressure drop, the allocation is then performed to the next level device, and the allocation strategy is that the high-priority device is allocated the pressure drop within its ability range, and the remaining target pressure drop is handed over to the subsequent device, until the overall pressure drop target is achieved or all devices run out of remaining space, i.e., the remaining down-regulation space of the device is exhausted, then the allocation is stopped; and the device regulation priority is written into the load pressure drop coordination database.

[0046] In the embodiment, through the sorting and allocation of the device regulation priority, fine decomposition of the load pressure drop instructions and efficient resource utilization are achieved. The process not only ensures that high-quality regulation resources respond to the pressure drop demand first, effectively improves the response speed and completion rate of the pressure drop regulation, but also checks the overall available regulation space, allocates tasks step by step and in stages, and prevents resource waste and instruction over-limit. At the same time, the related decision-making process and priority data are archived in the database throughout the whole process, which provides a solid support for subsequent regulation strategy optimization, pressure drop process tracing and self-learning evolution.

[0047] Specifically, the receiving adjustment priority list for pressure reduction adjustment, while combining the pretreated multi-dimensional regulation characteristic data and the regulation priority, carries out the specific process of pressure reduction adjustment decomposition among multiple power utilization equipment: receiving the current adjustment priority list, according to the real-time physical constraint and the total load pressure reduction target value, issuing the load pressure reduction instruction to each device, the physical constraint includes the device safe operation limit and the electrical parameter range, and the total pressure reduction target value is the pressure reduction adjustment demand quantity issued by the virtual power plant in this round; for each device, the multi-dimensional regulation characteristic data and the device regulation priority are obtained; the climbing ability is multiplied by the scheduling time step to obtain the dynamic limit allocation value, reflecting the maximum safe load adjustment quantity that can be achieved by the device in this round of scheduling period, ensuring that the allocation quantity does not exceed the physical ability of the device; at the same time, the device regulation priorities of all devices are summed to obtain the priority sum, the ratio of the device regulation priority of the device k to the priority sum is calculated, and multiplied by the total load pressure reduction target value to obtain the priority weighted allocation value, the total pressure reduction target is preliminarily allocated according to the ability proportion of each device, and the fairness and efficiency are considered; the remaining adjustable space of the device k, the dynamic limit allocation value and the priority weighted allocation value are compared, and the minimum value among the three is taken as the load pressure reduction allocation value, the three constraints ensure that the actual issued pressure reduction task of each device does not exceed its safe adjustable ability, and meets the global target and priority allocation principle, realizing the fine and robust load pressure reduction collaborative decomposition.

[0048] wherein the specific formula of the load pressure reduction allocation value is:

[0049] ;

[0050] In the formula, represents the load pressure reduction allocation value, which represents: the load pressure reduction value allocated to the device k in this round, and is subject to three constraints, and the minimum value is executed to prevent over-regulation and imbalance, and to realize safe, dynamic and optimal resource allocation; represents the remaining adjustable space of the device k, which prevents the allocation from exceeding the device limit, ensures the safety of the device, and avoids over-regulation; represents the climbing ability of the device k; represents the scheduling time step, which measures the time step of scheduling and control, that is, every how many time, the load pressure reduction allocation value is recalculated, feedback is detected, and new regulation action is issued; represents the dynamic limit allocation value, which prevents the scheduling allocation from being too fast and too fierce, avoids the damage to the device caused by transient impact, and realizes the regulation rate limiting; represents the device regulation priority; represents the total sum of the device regulation priorities of all adjustable devices; represents the total load pressure reduction target value; The priority weighted allocation value represents a proportion of the total pressure drop target allocated to the device according to the current device regulation priority, that is, the device with high priority, low risk and high ability undertakes more tasks, and the optimal utilization of resources is achieved.

[0051] In the embodiment, the intelligent decomposition and allocation of the multi-device load pressure drop regulation task are achieved by priority weighted allocation, dynamic physical constraint and joint control of the remaining space. The dynamic response limit, current adjustable space and overall pressure drop demand of the device itself are effectively considered, so that the pressure drop task allocated to each device is reasonable and safe, and the regulation resources are neither overloaded nor wasted. At the same time, the priority weighted mechanism fully utilizes the regulation ability of the high-priority device, improves the response efficiency and system resource utilization, and greatly improves the intelligence, adaptability and operation reliability of the building load pressure drop collaborative regulation system

[0052] Specifically, the specific process of feeding back the pressure drop regulation decomposition result is as follows: according to the calculated load pressure drop allocation value of each device, the specific regulation instruction of each device in this round is generated, the regulation instruction includes load down-regulation target, load pressure drop allocation value of each device, execution start and end time and related safety limit conditions, and is transmitted to each target device in real time through the controller; the state information of each device after regulation is collected and recorded in real time, including: actual power change, device running state, environmental parameters, electrical parameters and whether the device responds according to the instruction; the actual power change is used to quantify the load response effect, the device running state includes online, offline, fault and mode switching, the environmental parameters are temperature and humidity, lighting and air quality, the electrical parameters are current, voltage and harmonic content, and whether the device responds according to the instruction is determined by the state bit; the percentage deviation is used to compare the actual power change of the device with the load pressure drop allocation value, to determine whether the device completes the load pressure drop according to the plan, and to quickly identify the devices that do not complete the task and deviate from the plan; if the device response is abnormal, it is marked and fed back, the abnormal response includes not down-regulating according to the instruction, response lag, abnormal fluctuation and fault interruption, the abnormal device and the abnormal type are recorded and pushed to the dispatching or operation and maintenance end; the load pressure drop allocation value and the execution log in each regulation process are recorded, including the instruction issuing time, response feedback time and actual regulation effect, and are uploaded to the load pressure drop collaborative database.

[0053] In the embodiment, the device response abnormality can be identified, the execution effect and deviation in the regulation process can be dynamically recorded and marked, and the traceability and execution reliability of the regulation instruction are effectively improved. By storing the key data and execution log in the load pressure drop collaborative database, comprehensive and detailed historical data are provided for performance evaluation and operation and maintenance decision, a solid data foundation is laid for subsequent strategy optimization and intelligent self-learning, and the transparency, controllability and adaptability of the building load pressure drop collaborative regulation system are enhanced.

[0054] Specifically, based on the feedback from the voltage drop regulation decomposition results, the process of judging the overall coordinated regulation performance by integrating the voltage drop regulation decomposition results with the multi-dimensional regulation characteristic data before and after voltage drop regulation is as follows: The load voltage drop allocation value and actual regulation effect of each device are received. The load voltage drop allocation value is the target regulation amount issued by the command, and the actual regulation effect is the actual power change completed by the device. The deviation between the response effect of each device and the target is analyzed in real time to determine the overall voltage drop execution achievement. That is, by comparing the deviation between the target and the actual, the completion degree and response accuracy of this round of voltage drop regulation are quickly assessed. Based on all participants in the load voltage drop regulation... The total number of devices is determined, and the load voltage drop distribution values ​​of all devices are calculated and summed to obtain the total load voltage drop. The ratio of the total load voltage drop to the total target load voltage drop is subtracted by a constant, and the result is squared to obtain the voltage drop target achievement error. This error reflects the overall voltage drop execution effect; a smaller error indicates a higher degree of command completion, and is one of the core evaluation criteria for control performance. Electrical parameters of all devices before and after load voltage drop adjustment, as well as comfort scores before and after load voltage drop adjustment, are obtained from the load voltage drop coordination database. The voltage value before load voltage drop adjustment and the load voltage drop are calculated. The voltage change is obtained by taking the absolute value of the difference between the adjusted voltage values. This voltage change is then divided by the rated operating voltage constant and squared to obtain the relative voltage fluctuation value. The rated operating voltage constant is the standard voltage value marked on the equipment nameplate, such as 220V and 380V, used to normalize voltage disturbances across different equipment. The squaring process highlights extreme anomalies. The relative voltage fluctuation values ​​of all equipment are calculated and summed to obtain the total voltage fluctuation, quantifying the impact of this round of voltage drop regulation on the system's electrical stability. This is an important component of multi-objective risk assessment. Simultaneously, the standard deviation of the comfort score for the equipment's location within the historical period is calculated. The difference between the comfort score before and after load voltage drop adjustment is calculated, and the absolute value is taken to obtain the comfort change. The comfort change is divided by the standard deviation of the comfort score and squared to obtain the comfort fluctuation value. The comfort fluctuation values ​​of all equipment are calculated and summed to obtain the total comfort fluctuation. The voltage drop target achievement error, the total voltage fluctuation, and the total comfort fluctuation are added to obtain the multi-objective coordinated control performance evaluation value. The multi-objective coordinated control performance evaluation value comprehensively reflects the completion degree of voltage drop control, the impact of electrical disturbances and environmental comfort, and is the core criterion for multi-objective optimization and safety assurance.

[0055] The specific formula for the performance evaluation value of multi-objective coordinated regulation is as follows:

[0056] ;

[0057] In the formula, represents a multi-objective coordinated control performance evaluation value, used to measure the target completion degree, electrical disturbance risk and environmental comfort risk in the process of coordinated control of load pressure drop, is the core index of multi-objective coordinated optimization control effect evaluation in the scene of virtual power plant and building energy management, and the smaller the multi-objective coordinated control performance evaluation value is, the better the comprehensive performance of the current load pressure drop regulation is; represents a load pressure drop distribution value, reflecting the actual load pressure drop value shared and realized by the equipment in the current regulation, represents the load pressure drop total; represents a load pressure drop total target value, reflecting the current scheduling target, which is the amount of pressure drop task to be completed; represents a pressure drop target realization error, measuring the actual pressure drop completion in the current round, and the smaller the pressure drop target realization error is, the higher the pressure drop task completion degree is, and the constraint regulation action does not deviate from the pressure drop target, preventing under-regulation and over-regulation; represents a voltage change amount, indicating the absolute change amplitude of the voltage of the equipment in the regulation process, and quantifying the electrical disturbance caused by the equipment regulation; represents a rated working voltage constant, which is a constant, and is 220 and 380 according to different equipment types; represents a voltage fluctuation total, measuring the risk of voltage fluctuation relative to the rated value caused by the regulation of each equipment, and being normalized to facilitate horizontal comparison between different equipment; the smaller the voltage fluctuation total is, the lower the electrical disturbance is, and the safer and more stable the operation is; represents a comfort degree change amount, quantifying the change of environmental comfort after the current regulation, and measuring the influence on user experience; represents a comfort degree score standard deviation, reflecting the normal range of historical comfort degree fluctuation in the region, and being used for normalization of abnormal changes caused by regulation; represents a comfort degree fluctuation total, reflecting the ratio of comfort degree change caused by the regulation of each region to the historical normal fluctuation in the region, representing the abnormal risk of regulation, and the smaller the comfort degree fluctuation total is, the smaller the influence of regulation on the environment is, and the better the comfort is.

[0058] In the embodiment, by analyzing the key characteristic data before and after the pressure drop regulation, a multi-objective coordinated control performance evaluation mechanism integrating the pressure drop target completion degree, electrical disturbance risk and environmental comfort influence is established. The regulation effect can be judged in real time, comprehensively and quantitatively, which not only guarantees the accurate achievement of the pressure drop task, but also effectively monitors and suppresses the risks caused by electrical system fluctuation and user comfort decrease. The normalization of multiple evaluation indexes and the archiving analysis of historical standard deviation improve the abnormal detection and risk identification capability, provide a scientific data basis and criterion for subsequent regulation strategy optimization, adaptive adjustment and intelligent early warning, and greatly enhance the intelligence, adaptability and comprehensive operation safety of the building load pressure drop system.

[0059] Specifically, the specific process of identifying abnormal indicators and optimizing according to the overall coordinated regulation performance is as follows: after each load pressure drop regulation period ends, the multi-objective coordinated regulation performance evaluation value and three indicators are calculated, the three indicators include: pressure drop target error, voltage fluctuation total, and comfort fluctuation total, the three indicators respectively reflect the target achievement of this round of regulation task, the stability of the electrical system operation, and the change amplitude of the environmental comfort, which are important performance measurement parameters for multi-objective optimization; the change trend of the multi-objective coordinated regulation performance evaluation value and the three indicators is compared with the history, the history comparison includes comparison with the near cycle and extreme value interval statistics, which is used to timely find abnormal fluctuations and trend deviations, improve the sensitivity to abnormal conditions, identify abnormalities in pressure drop completion, electrical disturbance and comfort risk, abnormal judgment includes exceeding the historical normal interval, exceeding the abnormal disturbance threshold, continuous abnormal fluctuation, and realizes dynamic adaptive adjustment criterion; if the multi-objective coordinated regulation performance evaluation value is greater than the abnormal threshold and the three indicators are abnormal, a risk warning is issued, the abnormal equipment, time period and index item are recorded, the risk warning is pushed to the dispatching platform, and the equipment number, time period, over-limit index and abnormal type triggered by the alarm are marked in detail, which provides decision support for rapid response and processing; the multi-objective coordinated regulation performance evaluation value and the three indicators are written into the load pressure drop coordination database, and machine learning and rule engine are used for regular analysis, the machine learning method and the rule engine optimize the multi-objective coordinated regulation performance evaluation value algorithm and the discrimination threshold according to the historical data, realize the continuous self-learning and intelligent evolution of the algorithm, update the multi-objective coordinated regulation performance evaluation value algorithm, and periodically generate and archive the multi-objective coordinated regulation performance evaluation value trend report, abnormal index analysis, parameter adjustment record and optimization log.

[0060] In the embodiment, through continuous monitoring, historical comparison and dynamic alarm of the multi-objective coordinated regulation performance evaluation value and its sub-indicators, the identification and response of abnormal risks are realized. By archiving the abnormal information and detailed indicators, equipment, time period and other whole process, and combining machine learning and rule engine for regular analysis and self-optimization, the self-adaptive ability and abnormal disposal efficiency of the building load pressure drop regulation system to complex operation state are improved. At the same time, trend reports and optimization logs are generated, which provide data support and decision basis for continuous iteration, strategy adjustment and intelligent evolution.

[0061] Specifically, the specific process of continuously monitoring multi-dimensional regulation feature data, regulation priority, pressure drop regulation decomposition results and overall collaborative regulation performance, improving pressure drop distribution strategy and algorithm optimization is as follows: based on the real-time collected multi-dimensional regulation feature data, the environmental and electrical changes caused by load pressure drop regulation are monitored, including real-time tracking of the dynamic changes of environmental comfort score, voltage, current, harmonic content in the pressure drop regulation process, and timely discovery of potential risk hidden dangers in operation; periodically analyze the historical trends and abnormalities of the multi-objective collaborative regulation performance evaluation value and the three indexes, periodically analyze the core indexes of pressure drop target implementation error, voltage fluctuation total and comfort fluctuation total, identify fluctuation abnormalities and trend deviations by using statistical modeling and threshold discrimination methods, adjust the next round of load pressure drop regulation parameters according to the discovered abnormalities, if the comfort fluctuation total is continuously higher than the comfort fluctuation threshold, reduce the regulation amplitude of the equipment in the related area in the next regulation, that is, actively shrink the regulation amplitude, preferentially guarantee the environmental quality of the comfort sensitive area, and avoid affecting the user experience due to excessive pressure drop; if the voltage fluctuation total is higher than the voltage fluctuation threshold, limit the participation of equipment with voltage relative fluctuation value greater than the safety fluctuation threshold in pressure drop, by dynamically adjusting the participation range of the regulation equipment in real time, preferentially guarantee the safety and stability of the electrical system, and suppress the risk of electrical disturbance diffusion; at the same time, the feedback of user satisfaction and environmental parameters is fused, the user satisfaction is collected through regular investigation feedback by mobile APP, and the genetic algorithm is used to continuously optimize and self-learn the iteration of the device regulation priority and load pressure drop distribution value algorithm, wherein the genetic algorithm is a swarm intelligence optimization method, which simulates natural selection and genetic variation mechanism to adaptively evolve the device regulation priority parameters and load pressure drop distribution algorithm, continuously improving the adaptability and global optimality of the regulation strategy; through data-driven closed-loop feedback, the adaptive evolution of load pressure drop regulation and distribution strategy is realized, and the collaborative optimization closed loop of load pressure drop regulation from multi-source perception, intelligent decision, instruction execution to feedback optimization is constructed.

[0062] In the embodiment, multi-dimensional data of environment, electricity and regulation effect can be continuously monitored and analyzed to identify abnormal fluctuations and risk hidden dangers, and the pressure drop distribution parameters and regulation strategy are optimized in time. The self-learning ability of genetic algorithm is combined to continuously improve the intelligent level and global adaptability of the self-regulation strategy. The closed-loop mechanism effectively guarantees the environmental comfort, electrical safety and pressure drop task completion, and enhances the fine management ability and intelligent evolution ability of building load pressure drop collaborative regulation.

[0063] Reference Figure 2As shown, the second aspect of the present application provides a building load pressure drop collaborative optimization and distribution system for a virtual power plant, which is applied to the building load pressure drop collaborative optimization and distribution method for a virtual power plant, and includes: a multi-source data acquisition and environment monitoring module, which is used to obtain a scheduling time step, collect multi-dimensional regulation and control characteristic data in real time, construct a comfort evaluation model, and perform data preprocessing on the multi-dimensional regulation and control characteristic data; an adjustable resource identification and priority evaluation module, which is used to identify devices that can participate in load pressure drop regulation according to the multi-dimensional regulation and control characteristic data, evaluate the regulation priority of each device in combination with the preprocessed multi-dimensional regulation and control characteristic data and the comfort evaluation model, and generate a regulation priority list according to the regulation priority; a load pressure drop instruction decomposition and distribution module, which is used to receive the regulation priority list for pressure drop regulation, and perform pressure drop regulation decomposition among multiple power utilization devices in combination with the preprocessed multi-dimensional regulation and control characteristic data and the regulation priority, and feed back the pressure drop regulation decomposition result; a multi-target dynamic regulation and safety constraint module, which is used to identify abnormal indicators and perform optimization according to the feedback of the pressure drop regulation decomposition result, and judge the overall collaborative regulation performance by combining the pressure drop regulation decomposition result and the multi-dimensional regulation and control characteristic data before and after pressure drop regulation; and a user environment guarantee and optimization module, which is used to continuously monitor the multi-dimensional regulation and control characteristic data, the regulation priority, the pressure drop regulation decomposition result and the overall collaborative regulation performance, improve the pressure drop distribution strategy and perform algorithm optimization.

[0064] In the present embodiment, through the collaborative design of the multi-source real-time perception, intelligent priority evaluation, dynamic pressure drop distribution, multi-target collaborative regulation and user environment guarantee modules, the full-process automation and intelligentization of building load pressure drop regulation is realized. Not only can high-quality regulation resources be accurately identified and dynamically distributed, improving the scientificity and flexibility of pressure drop response, but also the safety and comfort risk in the regulation process can be evaluated in real time, and abnormalities can be found and optimized in a timely manner, ensuring operation safety and user experience. Through continuous adaptive monitoring and algorithm optimization, the self-learning and evolution ability of pressure drop regulation is improved, providing solid technical support and innovative value for virtual power plant demand response, building energy efficiency improvement and intelligent energy management.

[0065] It should be noted that, in this document, the terms such as first and second are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "contain" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or device.

[0066] The preferred embodiments of the application disclosed above are only to facilitate the understanding of the application. The preferred embodiments do not describe all the details necessary for the practice of the application and are not intended to limit the application to the particular embodiments described. As will be obvious to one of skill in the art, modifications and changes can be made without departing from the spirit and scope of the present application. The present description is chosen and described in order to best explain the principles of the application and its practical application to thereby enable others skilled in the art to best utilize the application. The application is limited only by the claims and their full scope and equivalents.

Claims

1. A method for collaborative optimization of building load voltage drop allocation for virtual power plants, characterized in that, Includes the following steps: S1, obtain the scheduling time step, collect multi-dimensional control feature data in real time, construct a comfort evaluation model, and perform data preprocessing on the multi-dimensional control feature data; S2. Based on the multidimensional control feature data, identify the equipment that can participate in load voltage drop regulation, and combine the pre-processed multidimensional control feature data with the comfort evaluation model to evaluate the control priority of each equipment, and generate a control priority list based on the control priority. The specific process of identifying equipment that can participate in load voltage drop regulation based on multidimensional regulation characteristic data, and evaluating the regulation priority of each equipment by combining the preprocessed multidimensional regulation characteristic data with the comfort evaluation model, is as follows: Acquire multi-dimensional control characteristic data to identify all devices currently eligible for load voltage drop regulation; for each device, calculate the ramp-up capability divided by the sum of the response time and a constant to obtain the response capability value; calculate the ratio of the remaining adjustable space to the current operating power to obtain the adjustable space ratio; acquire the harmonic content and voltage value within the historical period, calculate the mean and standard deviation of the harmonic content and the mean and standard deviation of the voltage value respectively, perform standard score normalization, and calculate the sum of the normalized values ​​of the harmonic content and voltage value as the electrical disturbance risk value; The equipment control priority is obtained by multiplying the response capability value, the adjustable space ratio, and the current comfort score, and then dividing by the sum of the electrical disturbance risk value and a constant. S3, receive the adjustment priority list to adjust the voltage drop, and at the same time combine the pre-processed multi-dimensional control feature data and control priority to decompose the voltage drop adjustment among multiple electrical devices, and provide feedback on the voltage drop adjustment decomposition results; S4. Based on the feedback of the pressure drop regulation decomposition results, the overall coordinated regulation performance is judged by combining the pressure drop regulation decomposition results with the multi-dimensional regulation characteristic data before and after pressure drop regulation. Abnormal indicators are identified and optimized based on the overall coordinated regulation performance. S5 continuously monitors multidimensional control characteristic data, control priority, voltage drop regulation decomposition results and overall collaborative control performance, improves voltage drop allocation strategy and performs algorithm optimization.

2. The building load voltage drop collaborative optimization allocation method for virtual power plants according to claim 1, characterized in that, The specific process of obtaining the scheduling time step, collecting multi-dimensional control feature data in real time, constructing a comfort evaluation model, and preprocessing the multi-dimensional control feature data is as follows: By deploying multiple types of sensors at key nodes in the building, real-time data on power load, electrical parameters, and environmental parameters are collected. The electrical parameters include harmonic content, voltage, and current values, while the environmental parameters include temperature, humidity, lighting brightness, and air quality in each area. The total target value of load voltage drop adjustment issued by the virtual power plant is recorded in real time. At the same time, the scheduling time step is obtained and defined as the time interval between each round of load voltage drop adjustment, allocation, and execution, serving as the time reference for each process. Standard load regulation test commands are issued to the equipment, and real-time operating data of the equipment is collected simultaneously. The response time and load change rate from the issuance of the command to the actual load change reaching the target are recorded, and the response speed of the equipment is calculated. At the same time, the controlled power regulation process of the equipment is analyzed, and the maximum rate of actual power change per unit time is calculated and screened as the ramp-up capability of the equipment. The rated minimum power constant of each piece of equipment is obtained, and the current operating power is collected in real time through field sensors. The remaining adjustable space is calculated using the difference between the current operating power and the rated minimum power constant. Based on real-time collected multi-dimensional environmental parameters, a multi-parameter environmental state dataset is constructed. Principal component analysis algorithm is used to perform clustering and dimensionality reduction analysis on the multi-parameter environmental state dataset to divide the natural distribution range of environmental parameters. The environmental state represented by each cluster center is taken as a typical comfort range. The principal component scores of each environmental parameter and the cluster center are used as environmental comfort reference values ​​and normalized to construct a comfort evaluation model and output the comfort scores corresponding to different environmental parameters. The total target value of load voltage drop, scheduling time step, power load data, electrical parameters, environmental parameters, response time, response speed, ramping ability, operating power, remaining adjustable space and comfort score are recorded as multi-dimensional control feature data. The multidimensional control feature data is time-aligned, standardized, and normalized. Median filtering is used to suppress noise. Outliers are removed by physical range verification, Hampel filtering, and sliding window standard score detection. Kalman filtering algorithm is used to imput missing data. A load drop collaborative database is constructed to store the multidimensional control feature data.

3. The building load voltage drop collaborative optimization allocation method for virtual power plants according to claim 1, characterized in that, The specific process for generating the adjustment priority list based on the adjustment priority is as follows: All devices that can participate in load voltage reduction regulation are sorted in descending order of device control priority to generate a control priority list; at the same time, combined with the total target value of load voltage reduction, the remaining downward adjustment space of all devices that can participate in load voltage reduction regulation is calculated to determine whether the overall voltage reduction capability meets the target requirements. Based on the priority list and the remaining downsizing space and ramping capacity of each device, voltage reduction instructions are assigned to each device sequentially. If the remaining adjustment space of the highest priority equipment is insufficient to cover the total load reduction target, it will be allocated to the next lower priority equipment until the overall load reduction target is achieved or all remaining space of all equipment is exhausted, at which point the allocation will stop; the equipment adjustment priority will be written into the load reduction coordination database.

4. The building load voltage drop collaborative optimization allocation method for virtual power plants according to claim 1, characterized in that, The specific process of decomposing voltage drop regulation among multiple electrical devices by using the receiving adjustment priority list and combining the preprocessed multi-dimensional control feature data and control priority is as follows: Receive the current adjustment priority list and issue load reduction instructions to each device according to real-time physical constraints and the total target value of load reduction; For each device, acquire multidimensional control feature data and device control priority; multiply the ramp-up capability by the scheduling time step to obtain the dynamic limit allocation value; simultaneously sum the device control priorities of all devices to obtain the priority sum, calculate the ratio of device control priority of device k to the priority sum, and multiply it by the total target value of load reduction to obtain the priority weighted allocation value; compare the remaining adjustable space of device k, the dynamic limit allocation value, and the priority weighted allocation value, and take the minimum value among the three as the load reduction allocation value.

5. The building load voltage drop collaborative optimization allocation method for virtual power plants according to claim 1, characterized in that, The specific process for feeding back the decomposition results of the pressure drop regulation is as follows: Based on the calculated load voltage drop distribution values ​​of each device, specific control instructions for each device in this round are generated and sent to each target device in real time through the controller; Real-time acquisition and recording of status information of each device after adjustment, including: actual power change, device operating status, environmental parameters, electrical parameters, and whether the device responds to instructions; By comparing the actual power change of the equipment with the load voltage drop distribution value, it is determined whether the equipment has completed the load voltage drop as planned. If an abnormal response of the equipment is found, it is marked and feedback is given. The load voltage drop distribution value and execution log of each round of adjustment process are recorded, including the time of instruction issuance, the time of response feedback and the actual adjustment effect, and are uploaded to the load voltage drop collaborative database.

6. The building load voltage drop collaborative optimization allocation method for virtual power plants according to claim 1, characterized in that, The specific process of judging the overall coordinated control performance based on the feedback of the pressure drop regulation decomposition results and the multi-dimensional control characteristic data before and after pressure drop regulation is as follows: Receive the load voltage drop distribution value and actual adjustment effect of each device, analyze the deviation between the response effect of each device and the target in real time, and determine the overall voltage drop execution status; Based on the total number of all devices involved in load drop regulation, calculate the load drop distribution value of all devices and sum them to obtain the total load drop. Subtract the ratio of the total load drop to the total target load drop value from the constant and square the result to obtain the error in achieving the load drop target. Obtain the electrical parameters of all equipment before and after load voltage drop adjustment, as well as the comfort scores before and after load voltage drop adjustment, from the load voltage drop coordination database; calculate the difference between the voltage value before and after load voltage drop adjustment and take the absolute value to obtain the voltage change; divide the voltage change by the rated operating voltage constant and square it to obtain the voltage relative fluctuation value; calculate the voltage relative fluctuation values ​​of all equipment and sum them to obtain the total voltage fluctuation; Simultaneously, the standard deviation of comfort scores in the area where the equipment is located is calculated within the historical period. The difference between the comfort scores before and after load pressure drop adjustment is calculated and the absolute value is taken to obtain the comfort change. The comfort change is divided by the standard deviation of comfort scores and squared to obtain the comfort fluctuation value. The comfort fluctuation values ​​of all equipment are calculated and summed to obtain the total comfort fluctuation. The sum of the voltage drop target achievement error, the voltage fluctuation, and the comfort fluctuation is added together to obtain the multi-objective coordinated regulation performance evaluation value.

7. The building load voltage drop collaborative optimization allocation method for virtual power plants according to claim 1, characterized in that, The specific process of identifying and optimizing abnormal indicators based on overall collaborative control performance is as follows: After each load voltage drop regulation cycle, the multi-objective coordinated control performance evaluation value and three indicators are calculated. The three indicators include: voltage drop target achievement error, total voltage fluctuation, and total comfort fluctuation. The changing trends of the multi-objective coordinated control performance evaluation value and the three indicators are compared with historical data to identify anomalies in voltage drop completion, electrical disturbance, and comfort risk. If the multi-objective coordinated control performance evaluation value is greater than the abnormal threshold or the three indicators are abnormal, a risk alarm is issued, and the abnormal equipment, time period, and indicator are recorded. The performance evaluation values ​​of multi-objective coordinated regulation and the three indicators are written into the load voltage drop coordination database. Machine learning and rule engine are used to analyze and update the multi-objective coordinated regulation performance evaluation value algorithm regularly. The multi-objective coordinated regulation performance evaluation value change trend report, abnormal indicator analysis, parameter adjustment record and optimization log are periodically generated and archived.

8. The building load voltage drop collaborative optimization allocation method for virtual power plants according to claim 1, characterized in that, The specific process of continuously monitoring multidimensional control feature data, control priority, voltage drop regulation decomposition results, and overall collaborative control performance, improving the voltage drop allocation strategy, and optimizing the algorithm is as follows: Based on real-time collected multi-dimensional control characteristic data, the system monitors environmental and electrical changes caused by load voltage drop regulation. It periodically analyzes the historical trends and anomalies of the multi-objective collaborative control performance evaluation values ​​and three indicators. For any anomalies detected, the system adjusts the load voltage drop regulation parameters for the next round. If the total comfort fluctuation consistently exceeds the comfort fluctuation threshold, the regulation amplitude of equipment in the relevant area is reduced in the next adjustment. If the total voltage fluctuation exceeds the voltage fluctuation threshold, equipment with a relative voltage fluctuation value greater than the safe fluctuation threshold is restricted from participating in voltage drop regulation. Simultaneously, by integrating user satisfaction and environmental parameter feedback, a genetic algorithm is used to continuously optimize and iterate the equipment control priority and load voltage drop allocation algorithm. Through data-driven closed-loop feedback, the system achieves adaptive evolution of load voltage drop regulation and allocation strategies, constructing a collaborative optimization closed loop for load voltage drop regulation.

9. A building load voltage drop collaborative optimization allocation system for virtual power plants, employing the building load voltage drop collaborative optimization allocation method for virtual power plants as described in any one of claims 1-8, characterized in that, include: The multi-source data acquisition and environmental monitoring module is used to obtain the scheduling time step, collect multi-dimensional control feature data in real time, construct a comfort evaluation model, and perform data preprocessing on the multi-dimensional control feature data. The adjustable resource identification and priority assessment module is used to identify equipment that can participate in load voltage drop regulation based on multi-dimensional regulation feature data, and to assess the regulation priority of each equipment by combining the pre-processed multi-dimensional regulation feature data and comfort evaluation model, and to generate a regulation priority list based on the regulation priority. The load voltage drop instruction decomposition and allocation module is used to receive the adjustment priority list for voltage drop adjustment, and at the same time, combine the pre-processed multi-dimensional control feature data and control priority to decompose the voltage drop adjustment among multiple electrical devices, and provide feedback on the voltage drop adjustment decomposition results. The multi-objective dynamic control and safety constraint module is used to judge the overall coordinated control performance based on the feedback of the pressure drop control decomposition results, the comprehensive pressure drop control decomposition results and the multi-dimensional control characteristic data before and after pressure drop control, and to identify abnormal indicators and optimize them based on the overall coordinated control performance. The user environment protection and optimization module is used to continuously monitor multi-dimensional control feature data, control priority, voltage drop regulation decomposition results and overall collaborative control performance, improve voltage drop allocation strategy and perform algorithm optimization.

Images