A method, device and storage medium for coordinated stability control of low voltage distribution network

Abstract

The present application discloses a method, device and storage medium for collaborative stability control of a low-voltage distribution network, and belongs to the technical field of energy storage control of power systems. By constructing a time-power characteristic model (P-t model) of power sources and loads based on historical data and predicted data, accurate identification of short-term high-power loads and long-term steady-state loads is achieved, and a collaborative optimization access method for supercapacitor and lithium battery energy storage is designed. A flexible collaborative switching strategy and a method combining PI control with repetitive control are adopted to achieve dynamic switching and efficient control of multi-energy storage systems, thereby improving the dynamic response speed and stability of the power grid. In addition, through terminal monitoring and multi-time scale data mining, combined with machine learning models to dynamically optimize energy storage scheduling strategies, the adaptability and economy of the energy storage system are improved, meeting the needs of complex load scenarios in new energy power systems, and providing technical support for the safe and efficient operation of low-voltage distribution networks.

Description

Technical Field

[0001] The embodiments of the present application relate to the technical field of power system energy storage control, and in particular to a method, device and storage medium for coordinated stability control of a low-voltage distribution network. Background Art

[0003] With the rapid development of new power systems characterized by a high proportion of new energy access, both the power supply and load sides of the power system show strong uncertainty. This uncertainty poses a huge challenge to the dispatching capacity of the power system and the safety and stability of the power grid.

[0004] At present, one of the more effective solutions is to adopt new energy storage technology with bidirectional regulation capabilities. Through the energy storage system, power supply and load can be effectively dispatched from multiple time and space perspectives, the balance of power grid flow can be achieved, thereby ensuring the safe and stable operation of the power grid.

[0005] In the low-voltage distribution network scenario of the new power system, with the increasing requirements of industrial and commercial parks for power quality and grid stability, the increasing demand for super-fast charging of electric vehicles, and the increasingly diversified demand for load regulation of smart grids, traditional single energy storage technology can no longer meet actual needs. This is mainly because traditional energy storage technology usually cannot simultaneously take into account performance requirements such as high power density, high energy density, fast response and long life. Therefore, it is particularly important to develop a multi-energy storage control system that can efficiently and compatible with multiple energy storage technologies. Summary of the invention

[0006] The present application provides a method, device and storage medium for coordinated stability control of a low-voltage distribution network. The technical solution is as follows:

[0007] On the one hand, a low-voltage distribution network collaborative stability control method is provided, the method is used for a new energy power system low-voltage distribution network, the method comprises:

[0008] Based on historical data and forecast data, the power characteristics of the application area are identified and the power Pt characteristic model is constructed;

[0009] Based on historical data and forecast data, the load power in the application area is characterized and a load Pt characteristic model is constructed, where the load characteristics determine the type of energy storage host;

[0010] Designing a supercapacitor and lithium battery energy storage access method according to the power source Pt characteristic model and the load Pt characteristic model, wherein the access method is designed for short-term high-power load and long-term steady-state load;

[0011] A multi-energy storage collaborative control strategy for dynamically switching between the supercapacitor and the lithium battery energy storage based on a flexible collaborative switching strategy;

[0012] Through the monitoring data of terminal electricity meters, grid connection points and distribution gateways, data mining and multi-time scale prediction of power supply and load power are carried out, and the multi-energy storage collaborative control strategy is dynamically adjusted in combination with the machine learning model.

[0013] Optionally, the method of identifying the characteristics of the power supply power in the application area based on the historical data and the predicted data and constructing a power supply Pt characteristic model includes:

[0014] Collect historical data of different power types and analyze forecast data for the power supply in the application area;

[0015] A power source Pt characteristic model with time and power as variables is constructed according to the historical data and the predicted data.

[0016] Optionally, constructing a power source Pt characteristic model with time and power as variables according to the historical data and the predicted data includes:

[0017] Acquire the historical data and the predicted data, and perform data screening and fitting to obtain a daily average output power curve;

[0018] According to the preset interval and the average power supply ratio, the grid factor rgrid[i] and the photovoltaic factor rpv[i] are determined, where the sum of the factors is 1, and i is the time series at the preset interval within a day;

[0019] The source Pt characteristic model is constructed according to the mains factor rgrid[i] and the photovoltaic factor rpv[i], which is recorded as Psource=Pgrid*rgrid[i]+Ppv*rpv[i], where Pgrid is the power supply of the grid and Ppv is the output power of the photovoltaic power generation;

[0020] Determine the time and power range to predict the power supply power in the current period;

[0021] Compare the power supply power in the current period with the real-time data to determine whether correction is required;

[0022] Feedback the final power supply Pt characteristic model.

[0023] Optionally, the method of identifying the characteristics of the load power in the application area based on the historical data and the forecast data and constructing a load Pt characteristic model includes:

[0024] Based on the historical and predicted data of the load in the application area, the characteristics of the load are identified to obtain short-term high-power load and long-term steady-state power load;

[0025] The load Pt characteristic model is constructed according to the characteristic identification result.

[0026] Optionally, constructing the load Pt characteristic model according to the characteristic identification result includes:

[0027] Acquire the historical data and the forecast data, perform data screening and fit the daily average load power curve;

[0028] According to the preset interval and the average power supply power ratio, the overcapacity factor rsc[i] and the lithium battery factor rlfp[i] are determined, where the sum of the factors is 1, and i is the time series at the preset interval within a day;

[0029] The load Pt characteristic model is constructed according to the supercapacitor factor rsc[i] and the lithium battery factor rlfp[i], which is recorded as Pload[x][i]=Psc*rsc[i]+Plf*rlfp[i], where Psc is the supercapacitor energy storage output power and Plf is the photolithium battery energy storage output power;

[0030] Determine the time and power range to predict the load power in the current period;

[0031] Compare the load power of the current period with the real-time data to determine whether the x-th load Pt characteristic model needs to be corrected;

[0032] Feedback the final load Pt characteristic model.

[0033] Optionally, the designing of a connection mode of supercapacitor and lithium battery energy storage according to the power source Pt characteristic model and the load Pt characteristic model includes:

[0034] A method of locally connecting a short-term high-power load to an overcapacity energy storage for compensatory discharge according to the power source Pt characteristic model and the load Pt characteristic model;

[0035] A method of accessing lithium battery energy storage for long-term steady-state loads according to the power source Pt characteristic model and the load Pt characteristic model.

[0036] Optionally, the multi-energy storage collaborative control strategy for dynamically switching between the supercapacitor and the lithium battery energy storage based on the flexible collaborative switching strategy includes:

[0037] Active load tracking is performed with load as the center, and a collaborative algorithm that uses load feature recognition to adaptively determine the host;

[0038] Dynamically switch the supercapacitor and the lithium battery energy storage through a flexible coordinated switching strategy of interval stepping and percentage gradual inclination;

[0039] A PI controller is selected to repeatedly control the dynamic switching.

[0040] On the other hand, a low-voltage distribution network collaborative stability control device is provided, the device is used for a low-voltage distribution network of a new energy power system, and the device includes:

[0041] The first module is used to identify the characteristics of the power supply in the application area based on historical data and predicted data and to construct a power supply Pt characteristic model;

[0042] The second module is used to identify the characteristics of the load power in the application area based on historical data and forecast data and build a load Pt characteristic model, where the load characteristics determine the type of energy storage host;

[0043] The third module is used to design an access mode of supercapacitor and lithium battery energy storage according to the power source Pt characteristic model and the load Pt characteristic model, wherein the access mode is designed for short-term high-power load and long-term steady-state load;

[0044] The fourth module is used for a multi-energy storage collaborative control strategy for dynamically switching between the supercapacitor and the lithium battery energy storage based on a flexible collaborative switching strategy;

[0045] The fifth module is used to perform data mining and multi-time scale prediction of power supply and load power through monitoring data from terminal power meters, grid connection points and distribution gateways, and dynamically adjust the multi-energy storage collaborative control strategy in combination with machine learning models.

[0046] On the other hand, a server is provided, comprising a processor and a memory; the memory stores at least one instruction, and the at least one instruction is used to be executed by the processor to implement the low-voltage distribution network collaborative stability control method as described in the above aspects.

[0047] On the other hand, a computer-readable storage medium is provided, wherein the storage medium stores at least one instruction, and the at least one instruction is used to be executed by a processor to implement the low-voltage distribution network collaborative stability control method as described in the above aspects.

[0048] On the other hand, a computer program product is also provided, which stores at least one instruction, and the at least one instruction is loaded and executed by the processor to implement the low-voltage distribution network collaborative stability control method described in the above aspect.

[0049] The present application provides a method for collaborative stability control of a low-voltage distribution network. By constructing a time-power characteristic model (Pt model) of power sources and loads based on historical data and predicted data, it realizes accurate identification of short-term high-power loads and long-term steady-state loads, and designs a collaborative optimization access method for supercapacitors and lithium battery energy storage. A flexible collaborative switching strategy and a method combining PI control with repetitive control are adopted to achieve dynamic switching and efficient control of multi-energy storage systems, thereby improving the dynamic response speed and stability of the power grid. In addition, through terminal monitoring and multi-time scale data mining, combined with machine learning models to dynamically optimize energy storage scheduling strategies, the adaptability and economy of the energy storage system are improved, meeting the needs of complex load scenarios in new energy power systems, and providing technical support for the safe and efficient operation of low-voltage distribution networks. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 A flow chart of a low voltage distribution network collaborative stability control method is shown;

[0051] Figure 2 A flow chart of a low voltage distribution network collaborative stability control method is shown;

[0052] Figure 3 A flow chart for power characteristic modeling of the power source Pt characteristic model and the load Pt characteristic model;

[0053] Figure 4 It is the flow chart of flexible cooperative switching algorithm;

[0054] Figure 5 shows a closed-loop control block diagram of repetitive control;

[0055] Figure 6 The block diagram of power closed-loop control based on the superposition of repetitive control and PI control is shown. DETAILED DESCRIPTION

[0056] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below in conjunction with the accompanying drawings.

[0057] The term "multiple" as used herein refers to two or more than two. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the related objects are in an "or" relationship.

[0058] Example 1

[0059] Please refer to Figure 1, which shows a flow chart of a low-voltage distribution network collaborative stability control method provided by an exemplary embodiment of the present application. The method is used for a low-voltage distribution network of a new energy power system, and the method includes:

[0060] Step 101 : Based on historical data and forecast data, characteristics of power sources in an application area are identified and a power source Pt characteristic model is constructed.

[0061] In a possible implementation, step 101 includes the following content.

[0062] Step S11 , collecting historical data of different power types and analyzing predicted data of power sources in the application area.

[0063] Step S12, constructing a power source Pt characteristic model with time and power as variables based on historical data and predicted data.

[0064] In one example, the historical data and forecast data of the power supply in the application area for 6 to 12 months were analyzed to obtain the output power of different power supplies in different time periods. The linear superposition method with influencing factors was used to construct the Pt characteristic model and database of the power supply, providing a basic basis for the regulation of energy storage installed at the power supply end, covering photovoltaic power generation, city power and other power supply types.

[0065] like Figure 3 As shown, the present application also provides a two-dimensional time-power characteristic modeling flow chart of the power source Pt characteristic model and the load Pt characteristic model.

[0066] Obtain historical data and forecast data, perform data screening and fit the daily average output power curve; determine the mains factor rgrid[i] and photovoltaic factor rpv[i] according to the preset interval and the average power supply power ratio, where the sum of each factor is 1, and i is the time series at the preset interval within a day; construct the source Pt characteristic model based on the mains factor rgrid[i] and the photovoltaic factor rpv[i], recorded as Psource=Pgrid*rgrid[i]+Ppv*rpv[i]; determine the time and power interval to predict the power supply power in the current period; compare the power supply power in the current period with the real-time data to determine whether it needs to be corrected; and feedback the final power supply Pt characteristic model. Where Pgrid is the power supply power of the grid, and Ppv is the output power of photovoltaic power generation.

[0067] Step 102, based on historical data and forecast data, characteristics of the load power in the application area are identified and a load Pt characteristic model is constructed, wherein the load characteristics determine the type of energy storage host.

[0068] In one example, the load in the application area is monitored and analyzed based on 6 to 12 months of historical data and forecast data, and the power of multiple types of loads in different time periods is obtained. The Pt characteristic model and database of each major load are constructed to distinguish between power-type and energy-type loads. In particular, short-term high-power loads are screened, such as fast charging, welding, smelting, lifting, motor starting, ejection and other similar loads.

[0069] In a possible implementation, step 102 includes the following content.

[0070] Step S21 , based on the historical data and forecast data of the load in the application area, characteristics of the load are identified to obtain a short-term high power load and a long-term steady-state power load.

[0071] Step S22: constructing a load Pt characteristic model according to the characteristic identification result.

[0072] like Figure 3 As shown, the present application also provides a two-dimensional time-power characteristic modeling flow chart of the power source Pt characteristic model and the load Pt characteristic model.

[0073] Obtain historical data and forecast data, perform data screening and fit the daily average load power curve; determine the supercapacitor factor rsc[i] and lithium battery factor rlfp[i] according to the preset interval and the average power supply power ratio, where the sum of each factor is 1, and i is the time series at the preset interval within a day; construct the load Pt characteristic model based on the supercapacitor factor rsc[i] and the lithium battery factor rlfp[i], recorded as Pload[x][i] = Psc*rsc[i] + Plf*rlfp[i]; determine the time and power interval to predict the load power of the current period; compare the load power of the current period with the real-time data to determine whether the xth load Pt characteristic model needs to be corrected; and feedback the final load Pt characteristic model. Where Psc is the output power of the supercapacitor energy storage, and Plf is the output power of the photolithium battery energy storage.

[0074] It can be seen that a major innovation of this application is that the related technology uses historical load data to judge load characteristics to regulate energy storage. In addition to collecting historical data on power supply and load power in energy storage scenarios such as industrial parks, the patent of this invention also combines the load spatial distribution to collect electric energy data in real time to perform basic source-load characteristic modeling; and uses online load characteristic tracking, load scenario identification, optimization of characteristic data and other self-learning methods to achieve power prediction. It not only has historical data characteristic models, but also real-time mathematics for model correction and use of predicted data for model optimization, laying the foundation for multi-dimensional energy storage layout and coordinated control. The modeling flow chart is as follows: Figure 3 As shown, different types of energy storage devices are matched with different load types. Figure 3In the correction coefficient, it is mainly obtained by looking up the table or fitting the quadratic curve, and it has the function of self-learning and correction. This patent recommends looking up the table, which has high execution efficiency. The premise that the two groups of influencing factors before and after the correction are added to 1 remains unchanged. When screening and identifying the short-term high-power load characteristic range, such as electric vehicle fast charging, rapid load regulation of virtual power plants, etc., such as the instantaneous high-power variation type in interval 1, the power and time laws of the required matching supercapacitor energy storage device are obtained. Both short-term and long-term can be designed and matched with lithium battery energy storage; the special interval 2 is matched with large-capacity supercapacitor energy storage.

[0075] Step 103, designing a connection method for supercapacitor and lithium battery energy storage according to the power source Pt characteristic model and the load Pt characteristic model, wherein the connection method is designed for short-term high-power load and long-term steady-state load.

[0076] Step S31, according to the power source Pt characteristic model and the load Pt characteristic model, a short-term high-power load is locally connected to the super-capacity energy storage for compensatory discharge.

[0077] Step S32, determining a lithium battery energy storage access method for a long-term steady-state load according to the power source Pt characteristic model and the load Pt characteristic model.

[0078] Based on the actual characteristics obtained in step 101 and step 102, a centralized energy storage access design scheme, a distributed energy storage access scheme or a hybrid energy storage access scheme can be designed in a targeted manner; short-term high-power loads can be connected to overcapacity energy storage for compensatory discharge, and long-term steady-state power loads can be controlled by lithium battery energy storage to achieve peak shaving and valley filling. The purpose of multi-energy storage design is to establish a basic model for coordinated and stable control of multi-energy storage.

[0079] Step 104, a multi-energy storage collaborative control strategy for dynamically switching between supercapacitor and lithium battery energy storage based on a flexible collaborative switching strategy.

[0080] In a possible implementation, step 104 includes the following content.

[0081] Step S41, actively tracking the load with the load as the center, and using load feature recognition to adaptively determine the host's collaborative algorithm.

[0082] Step S42, dynamically switch the supercapacitor and lithium battery energy storage through a flexible coordinated switching strategy of interval stepping and percentage gradual inclination.

[0083] Step S43, selecting a PI controller to repeatedly control the dynamic switching.

[0084] In one example, active load tracking is performed with the load as the center, the load characteristics in the multi-energy storage determine the type of energy storage host, and a collaborative algorithm that adaptively determines the host is adopted based on load characteristic recognition; in order to improve the dynamic stability of multi-energy storage control, a flexible and fast switching strategy of interval stepping and percentage gradual inclination is used to achieve dynamic switching between power-type and energy-type energy storage; in order to adapt to the accuracy of load power demand, stability of changes and rapidity, a controller based on PQ decoupling control uses a control algorithm that combines PI controller and repetitive control.

[0085] like Figure 4 Shown is the flow chart of the flexible cooperative switching algorithm.

[0086] In this application, ultra-capacity energy storage is mainly for short-term high-power loads, and lithium battery energy storage is mainly for long-term energy loads. The coordinated optimization control of the two types of energy storage makes up for the shortcomings of a single energy storage system and meets the multi-scenario requirements of the new power system. Figure 3 The requirements of application scenarios 1, 3, and 4, especially the second section, require an energy storage system that mainly uses high-power, large-capacity and ultra-capacity energy storage.

[0087] In order to coordinate the two energy storage devices, supercapacitor energy storage and lithium battery energy storage, especially the switching control, the patent of this invention innovatively proposes a flexible coordinated switching energy storage control strategy. Especially for short-term high-power demands (such as fast charging and ultra-fast charging of electric vehicles, welding, large motor starting, etc.), a flexible switching mode of dynamic load power matching method is proposed to ensure the response speed and smooth transition of the energy storage system; an adaptive host is designed in combination with the application scenario; in the process of switching between lithium battery and supercapacitor energy storage, the present invention introduces a 1-second flexible switching mechanism to avoid instantaneous fluctuations caused by hard switching. Different from the oscillation risk and poor system stability caused by the hard switching mode. Figure 4 As shown, for the load characteristic period of zone 1, the energy storage system EMS distributes the power of the two energy storage devices in 10 steps within 1 second with a slope of 1.

[0088] In addition, considering that the lithium battery energy storage system may have poor operating characteristics under extreme conditions such as high temperature or low temperature, the actual implementation process of the present invention can consider using overcapacity power supply to provide 20 to 30 degrees Celsius air conditioning or refrigerator power for the lithium battery energy storage. After stabilization, the overcapacity energy storage is removed, which can enhance the practicality of the application of the present invention and improve the system energy efficiency by about 0.7%.

[0089] Further, such as Figure 5 and 6 As shown, Figure 5 shows the closed-loop control block diagram of repetitive control, Figure 6 The block diagram of power closed-loop control based on the superposition of repetitive control and PI control is shown.

[0090] For the concept of multi-energy storage, typical application scenarios such as the coordinated control of dual energy storage types such as ultra-capacity energy storage and lithium battery energy storage, in addition to Figure 3 The source-charge characteristic modeling shown, Figure 4 In addition to the flexible switching control, it is also important to implement good steady-state performance and dynamic performance of the control. The present invention proposes a method combining PI control and repetitive control to ensure the accuracy of the entire energy storage power control, rapid dynamic response and stability, and adapt to the actual and widespread application scenarios of low-voltage distribution networks with harmonics in the mains.

[0091] The internal model principle of the mathematical basis of repetitive control is to add a delay recording link in the control path, record the error in each control cycle and feed it back to the error amount of the next cycle for superposition control, and achieve zero static error control of the system through periodic accumulation until the tracking error decays to zero. In actual design, it is easier to convert the delay link into a digital control method, so the internal model often appears in a discrete form.

[0092] Although repetitive control can ensure that the output waveform accurately tracks the given periodic reference signal, its dynamic response speed is slow, that is, there is a one-cycle delay in the error signal, which limits its application scenarios. PI control reacts quickly to the error signal, and the integrator can also correct the signal deviation in steady state. Therefore, the present invention adopts the "PI+repetitive control" superposition control algorithm. It is particularly noted that active and reactive power need to be decoupled and controlled. The present invention focuses on the control of active power P, and reactive power control is easier in practice and is not expanded here.

[0093] like Figure 6 As shown in FIG. 1 , it is a power closed-loop control block diagram of the compound controller, where N is the number of sampling times in one cycle, which is usually 200; Q is the internal model compensation coefficient, which is usually 0.95 to weaken the integral effect in the control process.

[0094] Due to the introduction of the integral link in the PI regulator, the fundamental component of the power grid can be quickly tracked without static error in the dq coordinate system, and the repetitive control can realize the tracking of multi-harmonic commands. Figure 6 The power control is mainly achieved through hierarchical control of energy storage EMS, coordination controller and PCS; according to the object model analysis of the present invention, in the PI controller, the ratio of P and I is preferably between 3 and 10 times, and the proportional coefficient Kp is first determined to be larger, such as 5, and then the integral coefficient Ki is determined to be 1.5.

[0095] Based on the above, the overall power control of the multi-energy storage system adopts the superposition of repeated control and PI control, which can be more accurate and reliable than the existing direct power EMS to PCS power control, and can also adapt to the actual widespread low-voltage distribution network application scenarios with harmonics in the mains.

[0096] Step 105, through the monitoring data of terminal power meters, grid connection points and distribution gateways, data mining and multi-time scale prediction of power supply and load power are performed, and the multi-energy storage collaborative control strategy is dynamically adjusted in combination with the machine learning model.

[0097] In one possible implementation, data monitoring and data mining at multiple spatial and temporal scales are mainly performed through electric energy meters at terminals, grid connection points, and distribution gateways, so as to predict power supply and load power and provide self-learning optimization functions for the next coordinated control of multi-energy storage.

[0098] In actual application scenarios, such as an industrial and commercial park, first collect and analyze power and load data according to the above steps to build a characteristic model and database. According to the load characteristics of the park, select distributed energy storage access, configure supercapacitors near the fast charging area, and deploy lithium batteries in long-term power consumption areas such as offices and production. During operation, the system follows the designed collaborative control strategy, tracks load changes in real time, dynamically allocates supercapacitors and lithium battery output, continuously optimizes subsequent control strategies based on electricity meter data, flexibly switches energy storage types to coordinate and control energy storage power, and improves system control accuracy and dynamic performance through the combination of EMS power repetitive control and PI control to ensure stable and reliable power supply to the park power grid.

[0099] In summary, the present invention systematically establishes a multi-energy storage collaborative stability control model for low-voltage distribution networks of new energy power systems, and proposes a collaborative control algorithm for energy storage scheduling of today's new distribution networks. Through the collaborative optimization control of supercapacitors (short-term high power) and lithium batteries (long-term energy type), it makes up for the shortcomings of a single energy storage system, and provides key technical support and application foundation for safe and reliable scheduling of distribution networks. Among them, the multi-energy storage system combines two or more energy storage technologies with different performance advantages to improve the overall performance and economy of the energy storage system, thereby meeting the multi-scenario segmentation needs of the new power system, and has core competitive advantages such as strong adaptability to working conditions, longer and safer energy storage life, better frequency regulation performance, and higher frequency regulation benefits.

[0100] Specifically, the technical effects of the low-voltage distribution network coordinated stability control scheme are mainly reflected in the following aspects:

[0101] Improve the stability and reliability of low-voltage distribution networks: By identifying the power characteristics of power sources and loads based on historical data and forecast data, a time-power characteristic model (Pt model) of power sources and loads is established, enabling the system to dynamically track and predict power output and load changes. Multi-energy storage collaborative control is achieved, and the dynamic collaboration of supercapacitors (high power density) and lithium batteries (high energy density) is adopted, so that short-term high-power loads and long-term steady-state loads can be matched with the optimal energy storage technology, significantly improving the dynamic response capability and load adaptability of the power grid.

[0102] Optimize the utilization efficiency of energy storage systems: Traditional energy storage systems often have difficulty meeting both high power density and high energy density requirements. This solution uses a flexible collaborative switching strategy to dynamically switch the operating modes of supercapacitors and lithium batteries. Supercapacitors are responsible for short-term high power requirements, such as fast charging of electric vehicles and welding, while lithium batteries are responsible for long-term energy requirements, such as peak shaving and valley filling and steady-state loads. This collaborative control method makes up for the shortcomings of a single energy storage technology, extends the service life of the energy storage device, and reduces the overall operation and maintenance costs of the system.

[0103] Improve the dynamic response speed and accuracy of the power grid: The flexible switching control strategy (based on interval stepping and percentage tilt) ensures the smooth transition of supercapacitors and lithium batteries during the switching process, avoiding the risk of instantaneous fluctuations and oscillations caused by hard switching. The closed-loop power control algorithm that combines PI control with repetitive control further enhances the energy storage system's ability to respond quickly to load fluctuations and its steady-state performance. It adapts to the complex working conditions of harmonics in low-voltage power distribution scenarios of the power grid, and achieves precise, zero-static-error power control.

[0104] Dynamic optimization based on machine learning: Real-time monitoring of power supply and load power is performed through terminal energy meters, grid connection points and distribution gateways, and multi-time scale data mining and prediction are used to dynamically adjust the energy storage collaborative control strategy. The machine learning model is used to optimize and iterate historical data and real-time data, which improves the accuracy of grid load power prediction and provides reliable data support for the next dispatch. This self-learning mechanism enables the system to continuously optimize and achieve rapid adaptation to different load scenarios.

[0105] Realize the economy and scenario adaptability of diversified energy storage: The system flexibly plans the energy storage access method according to different load scenarios (such as short-term high-power load, long-term steady-state load), including centralized, distributed or hybrid energy storage architecture. Reduce the waste of resources caused by improper configuration of traditional energy storage systems and improve the economy of energy storage systems. Especially for complex requirements in new energy scenarios (such as super-fast charging of electric vehicles, high-power industrial loads, etc.), this solution has demonstrated stronger scenario adaptability. Support multi-platform information interaction and resource integration: This solution can not only realize energy storage scheduling in low-voltage distribution networks, but also support efficient collaboration between different energy storage equipment and control strategies, providing technical support for the safe and reliable operation of distribution networks. At the information interaction level, this solution combines power sources, loads, and energy storage devices.

[0106] On the other hand, a low-voltage distribution network collaborative stability control device is provided, the device is used for a low-voltage distribution network of a new energy power system, and the device includes:

[0107] The first module is used to identify the characteristics of the power supply in the application area based on historical data and predicted data and to construct a power supply Pt characteristic model;

[0108] The second module is used to identify the characteristics of the load power in the application area based on historical data and forecast data and build a load Pt characteristic model, where the load characteristics determine the type of energy storage host;

[0109] The third module is used to design an access mode of supercapacitor and lithium battery energy storage according to the power source Pt characteristic model and the load Pt characteristic model, wherein the access mode is designed for short-term high-power load and long-term steady-state load;

[0110] The fourth module is used for a multi-energy storage collaborative control strategy for dynamically switching between the supercapacitor and the lithium battery energy storage based on a flexible collaborative switching strategy;

[0111] The fifth module is used to perform data mining and multi-time scale prediction of power supply and load power through monitoring data from terminal power meters, grid connection points and distribution gateways, and dynamically adjust the multi-energy storage collaborative control strategy in combination with machine learning models.

[0112] An embodiment of the present application also provides a computer-readable storage medium, in which at least one instruction is stored. The at least one instruction is loaded and executed by a processor to implement the low-voltage distribution network collaborative stability control method provided in the above embodiments.

[0113] Optionally, the computer readable storage medium may include: a read-only memory (ROM), a random access memory (RAM), a solid state drive (SSD), or an optical disk, etc. Among them, the random access memory may include a resistance random access memory (ReRAM) and a dynamic random access memory (DRAM).

[0114] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0115] A person skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware or by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.

[0116] The above description is only an optional embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A low voltage distribution network coordinated stability control method, characterized in that: The method is used for a low-voltage distribution network of a new energy power system, and the method comprises: Based on historical data and forecast data, the power characteristics of the application area are identified and the power Pt characteristic model is constructed; Based on historical data and forecast data, the load power in the application area is characterized and a load Pt characteristic model is constructed, where the load characteristics determine the type of energy storage host; Designing a supercapacitor and lithium battery energy storage access method according to the power source Pt characteristic model and the load Pt characteristic model, wherein the access method is designed for short-term high-power load and long-term steady-state load; A multi-energy storage collaborative control strategy for dynamically switching between the supercapacitor and the lithium battery energy storage based on a flexible collaborative switching strategy; Through the monitoring data of terminal electricity meters, grid connection points and distribution gateways, data mining and multi-time scale prediction of power supply and load power are carried out, and the multi-energy storage collaborative control strategy is dynamically adjusted in combination with the machine learning model.

2. The method according to claim 1, characterized in that The method of identifying the characteristics of the power supply power in the application area based on the historical data and the predicted data and constructing the power supply Pt characteristic model includes: Collect historical data of different power types and analyze forecast data for power supply in the application area; A power source Pt characteristic model with time and power as variables is constructed according to the historical data and the predicted data.

3. The method according to claim 2, characterized in that The constructing of a power source Pt characteristic model with time and power as variables according to the historical data and the predicted data comprises: Acquire the historical data and the predicted data, and perform data screening and fitting to obtain a daily average output power curve; According to the preset interval and the average power supply ratio, the grid factor rgrid[i] and the photovoltaic factor rpv[i] are determined, where the sum of the factors is 1, and i is the time series at the preset interval within a day; The source Pt characteristic model is constructed according to the mains factor rgrid[i] and the photovoltaic factor rpv[i], which is recorded as Psource=Pgrid*rgrid[i]+Ppv*rpv[i], where Pgrid is the power supply of the grid and Ppv is the output power of the photovoltaic power generation; Determine the time and power range to predict the power supply power in the current period; Compare the power supply power in the current period with the real-time data to determine whether correction is required; Feedback the final power supply Pt characteristic model.

4. The method according to claim 1, characterized in that: The method of identifying the characteristics of the load power in the application area based on the historical data and the predicted data and constructing the load Pt characteristic model includes: Based on the historical and predicted data of the load in the application area, the load characteristics are identified to obtain short-term high-power load and long-term steady-state power load; The load Pt characteristic model is constructed according to the characteristic identification result.

5. The method according to claim 4, characterized in that The step of constructing the load Pt characteristic model according to the characteristic identification result includes: Acquire the historical data and the forecast data, perform data screening and fit the daily average load power curve; According to the preset interval and the average power supply power ratio, the overcapacity factor rsc[i] and the lithium battery factor rlfp[i] are determined, where the sum of the factors is 1, and i is the time series at the preset interval within a day; The load Pt characteristic model is constructed according to the supercapacitor factor rsc[i] and the lithium battery factor rlfp[i], which is recorded as Pload[x][i]=Psc*rsc[i]+Plf*rlfp[i], where Psc is the supercapacitor energy storage output power and Plf is the photolithium battery energy storage output power; Determine the time and power range to predict the load power in the current period; Compare the load power of the current period with the real-time data to determine whether the x-th load Pt characteristic model needs to be corrected; Feedback the final load Pt characteristic model.

6. The method according to claim 1, characterized in that The design of the access mode of supercapacitor and lithium battery energy storage according to the power source Pt characteristic model and the load Pt characteristic model includes: A method of locally connecting a short-term high-power load to an overcapacity energy storage for compensatory discharge according to the power source Pt characteristic model and the load Pt characteristic model; A method of accessing lithium battery energy storage for long-term steady-state loads according to the power source Pt characteristic model and the load Pt characteristic model.

7. The method according to claim 1, characterized in that The multi-energy storage collaborative control strategy for dynamically switching between the supercapacitor and the lithium battery energy storage based on the flexible collaborative switching strategy includes: Active load tracking is performed with load as the center, and a collaborative algorithm that uses load feature recognition to adaptively determine the host; Dynamically switch the supercapacitor and the lithium battery energy storage through a flexible coordinated switching strategy of interval stepping and percentage gradual inclination; A PI controller is selected to repeatedly control the dynamic switching.

8. A low voltage distribution network coordinated stability control device, characterized in that: The device is used for a low-voltage distribution network of a new energy power system, and the device comprises: The first module is used to identify the characteristics of the power supply in the application area based on historical data and predicted data and to construct a power supply Pt characteristic model; The second module is used to identify the characteristics of the load power in the application area based on historical data and forecast data and build a load Pt characteristic model, where the load characteristics determine the type of energy storage host; The third module is used to design an access mode of supercapacitor and lithium battery energy storage according to the power source Pt characteristic model and the load Pt characteristic model, wherein the access mode is designed for short-term high-power load and long-term steady-state load; The fourth module is used for a multi-energy storage collaborative control strategy for dynamically switching between the supercapacitor and the lithium battery energy storage based on a flexible collaborative switching strategy; The fifth module is used to perform data mining and multi-time scale prediction of power supply and load power through monitoring data from terminal power meters, grid connection points and distribution gateways, and dynamically adjust the multi-energy storage collaborative control strategy in combination with machine learning models.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and the computer program is loaded and executed by a processor to implement the low-voltage distribution network collaborative stability control method as described in any one of claims 1 to 7 above.

Images