Self-adaptive scheduling decision energy management system

Abstract

The self-adaptive scheduling decision energy management system comprises an energy tracking unit (ETU), an energy prediction unit (EPU), a system decision unit (SDU) and a cooperative control unit (COCU). According to the invention, basic information of other subsystem units is collected, and according to photovoltaic power generation conditions, power grid conditions, actual use conditions of loads and customer expectation strategies, energy storage charging and discharging and diesel engine starting and stopping strategies are dynamically adjusted while photovoltaic treatment is well done, power supply quality is guaranteed and green power is fully used, so that optimal system performance is realized, and the system performance is improved. The charging and discharging safety of the battery cabinet is ensured, the charging and discharging times are reduced, the comprehensive cost is lowest, and the benefit is highest.

Description

Technical Field

[0001] The present invention relates to the field of photovoltaic energy storage, and in particular to an adaptive scheduling decision-making energy management system. Background Art

[0002] In a photovoltaic storage system, photovoltaics are usually connected to the DC side of the energy storage grid. Since photovoltaic power and load are unstable, an energy management system is needed to manage photovoltaics and the convergence unit so that the entire system can fully utilize photovoltaic green electricity under the premise of peak shaving and valley filling and time-sharing charging and discharging.

[0003] In order for the system to truly not waste light and give full play to the maximum role of photovoltaics, it is necessary to make full use of photovoltaics during the discharge period without over-discharging the battery; during the charging process, when the battery is close to full, ensure that the photovoltaics continue to output full power without overcharging the battery. While achieving this effect, it is necessary to solve the problem of battery back-and-forth charging and discharging oscillations during the charging and discharging process, reduce the number of charging and discharging times, and extend the battery life.

[0004] In order to prevent the overcharge and over-discharge of batteries caused by the oscillation of photovoltaic and load power during the charging and discharging process, the solution of most systems on the market is that when the SOC of the energy storage cabinet approaches the control point, the energy dispatching system will directly reduce or disable photovoltaic power, and use the stability of the power output of the power grid to ensure the safety of battery cell charging and discharging, but it will cause a large loss of photovoltaic green electricity.

[0005] Therefore, it is necessary to provide an adaptive scheduling decision energy management system to solve the above problems. Summary of the invention

[0006] The purpose of the present invention is to provide an adaptive scheduling decision energy management system.

[0007] The present invention achieves the above object through the following technical solutions:

[0008] An adaptive scheduling decision-making energy management system includes an energy tracking unit (ETU), an energy prediction unit (EPU), a system decision unit (SDU), a collaborative control unit (COCU), and other submodules; wherein the energy tracking unit uses different time intervals for different devices to monitor the subsystem equipment, confirm the online and offline status of the equipment, and obtain the subsystem equipment status data; at the same time, through data governance, different data are screened and filtered, stored in the database, and analyzed by the energy prediction unit; the energy prediction unit mainly collects the data of the current module, completes the prediction of the data of each module, and provides appropriate results to the system decision unit for strategy regulation; the short-term prediction is based on the statistics of photovoltaic data in different regions, and obtains the longest sunshine time Y according to the regional dimension information and date. pv; Photovoltaic changes are directly related to light, which is related to factors such as longitude and latitude (lat, lng), date (date), air temperature and humidity (T, H), wind speed (Vw), and wind direction (WD). Based on longitude and latitude and date information, the theoretical light time T is calculated. These weather factors can be used as input features to calculate the light factor (WF). The photovoltaic power is P, and the photovoltaic power generation G pv =WF×P×T; using the following formula: G=Y pv ×β+∈ t Solve the linear regression coefficient to predict the optimal photovoltaic power generation of the system; use the sliding average method, taking 3 days as a cycle, to predict the average daily photovoltaic power generation for the next 7 days. Power generation on that day: G x =(G x-3 +G x-2 +G x-1 ) / 3; by fitting the data of the above two algorithms, the predicted data of photovoltaic power generation is obtained as the data input of the decision-making unit; the system decision-making unit optimizes the control recommendation strategy of the decision-making system based on the basic data related to the system.

[0009] Furthermore, the energy tracking unit also screens and filters different data through data management, stores them in the database, and provides them to the energy prediction unit for analysis.

[0010] Furthermore, the energy prediction unit mainly collects data including photovoltaic power generation, load power, energy storage SOC, load power, etc.

[0011] Furthermore, the load power consumption is predicted, the phase voltage and power of the load are counted, the short-term high load peak characteristics are obtained, and the peak time, duration, and power value information are recorded to form a feature library.

[0012] Furthermore, the system decision unit is based on system-related basic data, including: photovoltaic power generation trend, power supply trend, and load power consumption trend.

[0013] Compared with the prior art, the present invention collects basic information of other subsystem units, targets photovoltaic power generation, power grid conditions, actual load usage, and customer expectation strategies, and while doing a good job in photovoltaic management, ensuring power supply quality, and making full use of green electricity, dynamically adjusts the energy storage charging and discharging and diesel generator start and stop strategies to achieve optimal system performance, ensure the safety of battery cabinet charging and discharging, and reduce the number of charging and discharging times, thus achieving a control strategy with the lowest overall cost and the highest benefit. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 It is a schematic diagram of the present invention. DETAILED DESCRIPTION

[0015] Example:

[0016] See also Figure 1 The entire photovoltaic storage system consists of grid (GRID), load (Load), energy management system (EMSU), battery management system (BMS), photovoltaic (PV), energy storage inverter (PCS), static transfer switch (STS), automatic power transfer switch (ATS), generator (Generator) and other module units.

[0017] This embodiment demonstrates an adaptive scheduling decision energy management system, the energy management system includes an energy tracking unit (ETU), an energy prediction unit (EPU), a system decision unit (SDU), and a collaborative control unit (COCU).

[0018] Energy tracking unit: The energy tracking unit uses different time intervals for different devices to monitor subsystem devices, confirm the online and offline status of devices, and obtain subsystem device status data. At the same time, through data management, different data are filtered and stored in the database for analysis by the energy prediction unit.

[0019] Energy prediction unit, the energy prediction unit mainly collects data of the current module, including photovoltaic power generation, load power, energy storage SOC, load power, etc., completes the prediction of data of each module, and provides appropriate results to the system decision unit for strategy control.

[0020] The short-term forecast is based on the statistics of photovoltaic data in different regions, and the longest sunshine time Y is obtained according to the regional dimension information and date. pv The change of photovoltaic power is directly related to the light, which is related to factors such as longitude and latitude (lat, lng), date (date), air temperature and humidity (T, H), wind speed (Vw), and wind direction (WD). Based on the longitude and latitude and date information, the theoretical light time T is calculated. These weather factors can be used as input features to calculate the light factor (WF). The photovoltaic power is P, and the photovoltaic power generation G pv =WF×P×T. Use the following formula: G=Y pv ×β+∈ t Solve for the linear regression coefficients and predict the optimal photovoltaic power generation for the system.

[0021] The sliding average method is used to predict the average daily photovoltaic power generation for the next 7 days, with 3 days as a cycle. Power generation for the day: G x =(G x-3 +G x-2 +G x-1 ) / 3.

[0022] By fitting the data of the above two algorithms, the predicted data of photovoltaic power generation is obtained as the data input of the decision-making unit.

[0023] Predict load power consumption, count load phase voltage and power, obtain short-term high load peak characteristics, and record peak time, duration, and power value information to form a feature library. Use energy storage and photovoltaics to manage load peaks, reduce grid load peaks, and improve grid load balance.

[0024] Make a systematic forecast of the power supply of the power grid, start energy storage in advance according to the characteristics of power instability (feed voltage value, duration, occurrence time, etc.), respond to the power grid demand, adjust the total output power, and ensure the stable output of electricity.

[0025] The system decision unit is based on system-related basic data, including: photovoltaic power generation trend, power supply trend, load power consumption trend. Through algorithm analysis, it optimizes the control recommendation strategy of the decision system to fully absorb green electricity and obtain the best benefits.

[0026] In a system where the photovoltaic power is greater than the energy storage cabinet power, during the PCS discharge process, when the energy storage cabinet SOC is small and the energy storage needs to be charged, the photovoltaic fluctuation will cause the energy storage charging and discharging to oscillate, resulting in an increase in the number of battery cell charge and discharge times. The decision-making system conducts comprehensive analysis and evaluation of multi-dimensional data to effectively increase the number of charge and discharge times within the safe use range of the battery and increase the battery cell life.

[0027] For example, the data collected by the energy management unit can be analyzed, and the available power of the energy storage cabinet, the state of the grid, and the state of the DC source can be obtained to control the start and stop of the diesel generator, so as to achieve continuous power supply to the grid at the lowest cost. According to different lighting conditions and real-time status, different power sources can be used to charge the energy storage cabinet to achieve full consumption of green electricity.

[0028] Collaborative control unit, the design of the energy storage system needs to fully consider the overall safety and stability. Once an accident occurs, the impact is extremely significant. Therefore, the overall system needs to ensure safe and stable operation under various conditions. In addition to the fire protection system, early warning and emergency response system of the cabinet itself, the energy storage system also fully considers the extreme situation that when the main control is abnormal, other functional modules continue to work, which will cause the system to lose control and affect the safety of system use. In addition to the original EMS management system, a collaborative control unit is added. When the main system module is abnormal, it can monitor the module abnormality in real time, and implement different safety measures according to different response strategies to ensure system safety.

[0029] The collaborative control unit is an external system of the energy management system. Adding a set of redundant protection equipment can reduce the occurrence of abnormal situations such as overcharging and over-discharging, making the system safer. It mainly undertakes the following functions: 1) High-frequency monitoring of system equipment status to obtain real-time module status data. 2) Through data analysis, timely determine whether there are abnormalities in the system module; 3) After discovering the abnormality of the main system module, issue instructions to adjust the equipment parameters, or even make the system stop and power off to prevent secondary abnormalities in the system equipment.

[0030] Compared with the existing technology, by collecting basic information of other subsystem units, targeting photovoltaic power generation, power grid conditions, actual load usage, and customer expectation strategies, while doing a good job in photovoltaic management, ensuring power supply quality, and making full use of green electricity, it dynamically adjusts the energy storage charging and discharging and diesel generator start and stop strategies to achieve optimal system performance, ensure the safety of battery cabinet charging and discharging, and reduce the number of charging and discharging times, thus achieving the lowest overall cost and highest efficiency control strategy.

[0031] The above are only some embodiments of the present invention. For those skilled in the art, several modifications and improvements can be made without departing from the creative concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. An adaptive scheduling decision energy management system, characterized by: It includes multiple sub-modules, such as energy tracking unit (ETU), energy prediction unit (EPU), system decision unit (SDU), collaborative control unit (COCU), etc. Among them, the energy tracking unit uses different time intervals for different devices to monitor the subsystem equipment, confirm the online and offline status of the equipment, and obtain the subsystem equipment status data; at the same time, through data governance, different data are filtered and stored in the database for analysis by the energy prediction unit; the energy prediction unit mainly collects the data of the current module, completes the prediction of the data of each module, and provides appropriate results to the system decision unit for strategy regulation; the short-term prediction is based on the statistics of photovoltaic data in different regions, and obtains the longest sunshine time Y according to the regional dimension information and date. pv ; Photovoltaic changes are directly related to light, which is related to factors such as longitude and latitude (lat, lng), date (date), air temperature and humidity (T, H), wind speed (Vw), and wind direction (WD). Based on longitude and latitude and date information, the theoretical light time T is calculated. These weather factors can be used as input features to calculate the light factor (WF). The photovoltaic power is P, and the photovoltaic power generation G pv =WF×P×T; using the following formula: G=Y pv ×β+∈ t Solve the linear regression coefficient to predict the optimal photovoltaic power generation of the system; use the sliding average method, taking 3 days as a cycle, to predict the average daily photovoltaic power generation for the next 7 days. Power generation on that day: G x =(G x-3 +G x-2 +G x-1 ) / 3; by fitting the data of the above two algorithms, the predicted data of photovoltaic power generation is obtained as the data input of the decision-making unit; the system decision-making unit optimizes the control recommendation strategy of the decision-making system based on the basic data related to the system.

2. The adaptive scheduling decision energy management system according to claim 1, characterized in that: At the same time, the energy tracking unit uses data management to screen and filter different data, store them in the database, and provide them to the energy prediction unit for analysis.

3. The adaptive scheduling decision energy management system according to claim 1, characterized in that: The energy prediction unit mainly collects data including photovoltaic power generation, load power, energy storage SOC, load power, etc.

4. The adaptive scheduling decision energy management system according to claim 1, characterized in that: Predict load power consumption, count the phase voltage and power of the load, obtain short-term high load peak characteristics, and record the peak time, duration, and power value information to form a feature library.

5. The adaptive scheduling decision energy management system according to claim 1, characterized in that: The system decision unit is based on basic system-related data, including: photovoltaic power generation trend, power supply trend, and load power consumption trend.

Images