A micro energy storage voltage active regulation method for long-term support of low voltage of a rural power grid

Abstract

A micro energy storage voltage active regulation method for long-term support of low voltage in rural power grids, comprising: based on the state monitoring of the power grid voltage of the low-voltage frequent area, the process modeling of the state of the power grid and the adjustment action according to the series micro energy storage voltage regulation characteristics; taking the long-term voltage support capability as the target, a micro energy storage system voltage regulation value evaluation model is constructed; based on the constructed state, action and value model, the voltage regulation of the micro energy storage system is trained through deep deterministic policy gradient optimization; for the power grid area information under fluctuation, the working mode and target voltage of the series micro energy storage system are gradually set according to the trained model. The present application can provide long-term voltage support capability for the low voltage phenomenon of the remote power grid area in rural areas under the condition of sudden increase of load, effectively enhance the intelligent level of the active regulation of the battery of the micro energy storage system, and has important value for improving the voltage quality management ability of the rural power grid.

Description

Technical Field

[0001] This invention belongs to the field of energy storage technology application, specifically a micro-energy storage voltage and active power regulation method for long-term low-voltage support of rural power grids. Background Technology

[0002] Currently, the rapid growth in load types and electricity consumption in rural areas has led to frequent low-voltage phenomena in the power grid. Because rural areas are often remote, distribution substations frequently have numerous long, single-radial lines, posing challenges to power reliability and quality. Compared to urban residents and industrial users, rural electricity loads exhibit stronger seasonal characteristics, such as peak summer demand and sudden load surges during peak farming seasons, resulting in low voltage. Therefore, large-scale grid upgrades are not an economical option for these areas. Similarly, considering the fluctuating nature of voltage usage, methods such as on-load tap changers or voltage regulators are insufficient for accurately configuring regulation and cannot provide immediate responses to voltage regulation needs.

[0003] To address the aforementioned needs, this invention provides a low-voltage regulation and mitigation solution for rural power grids using micro-energy storage systems. However, micro-energy storage systems often have limited battery capacity, and disordered charging and discharging not only affect battery life but also weaken the system's ability to support grid voltage. Therefore, it is urgent to develop a voltage regulation method for micro-energy storage systems that combines the dynamic characteristics of rural power grid voltage and addresses the long-term goal of supporting low-voltage phenomena. Summary of the Invention

[0004] This invention provides a micro-energy storage voltage active power regulation method for long-term support of low voltage in rural power grids. The technical problem it solves is the low voltage phenomenon that occurs in remote rural power grid areas under sudden load increases. To achieve the goal of long-term voltage support for the power grid, a process model for power grid voltage regulation is constructed, and an intelligent voltage regulation strategy for series micro-energy storage systems is given based on deep deterministic strategy gradient learning.

[0005] The specific technical solution of the present invention is as follows:

[0006] A method for voltage and active power regulation of micro-energy storage for long-term low-voltage support in rural power grids includes the following steps:

[0007] Step 1: Monitor the voltage status of the power grid in areas with frequent low voltage, and based on the voltage regulation characteristics of series micro-energy storage, perform process modeling of the power grid status and regulation actions to provide an information expression model for regulation strategies;

[0008] Step 2: Construct a voltage regulation value assessment model for micro energy storage systems with the goal of long-term voltage support capability;

[0009] Step 3: Based on the grid state, regulation actions, and voltage regulation value assessment model of the micro-energy storage system constructed in Step 1 and Step 2, construct a training network, and train the voltage regulation strategy of the micro-energy storage system through deep deterministic strategy gradient optimization to obtain the training model.

[0010] Step 4: Based on the training model in Step 3 and the actual grid distribution area status information, the working mode and target voltage of the series micro energy storage system are gradually set to achieve active power regulation of the micro energy storage voltage in the actual operating environment.

[0011] Furthermore, in step 1, based on the voltage regulation characteristics of series micro-energy storage, process modeling of the grid state and regulation actions is performed, specifically including:

[0012] The voltage, current, and power values ​​of the micro-energy storage system's installation nodes are acquired through sensor monitoring devices and denoted as U. O I O P O and Q O Based on the following electrical connection formula:

[0013]

[0014] Q O =I O 2 X u

[0015] P O =I O 2 R u +P ESS

[0016] The system resistance R is calculated. u Reactance X u , where P ESS The active power of the micro-energy storage system is positive when charging and negative when discharging. Based on this, a process model of voltage regulation of the micro-energy storage system is constructed, where the grid state model is... s t This represents the power grid state information at time t, including system voltage. Energy storage battery capacity e t Reactance z t and time characteristics δ t ;

[0017] Considering the regulation rules of the micro-energy storage system, the action model is denoted as: m t The operating modes are categorized into three types: charging, discharging, and standby. The target supporting voltage value for the energy storage system;

[0018] Among them, a voltage regulation function is designed to characterize the parameter relationship between the power grid state and the regulation process, denoted as . When the battery capacity meets the requirements

[0019]

[0020] s t (e t ) indicates with The remaining capacity of the energy storage battery after adjusting to the target; when the battery capacity is insufficient to meet the target voltage support requirements, the actual adjustment situation is calculated according to the following formula:

[0021]

[0022] This represents the maximum allowable charging / discharging power of the energy storage at time t.

[0023] Furthermore, the voltage regulation value assessment model for the micro-energy storage system in step 2 separately evaluates the grid voltage support utility and the remaining battery energy, and provides the final evaluation value of the micro-energy storage regulation action in a weighted summation form, providing a basis for network model training for subsequent regulation strategies. Its expression is denoted as:

[0024]

[0025] Where ω1 and ω2 are weight parameters, set empirically, r u (·), r e (·) represent the value of the grid voltage after regulation by the micro-energy storage system and the future supporting value of the remaining battery capacity, respectively; where r u (·)~s v (u), r u (·)~s t (u);

[0026] Set the upper limit U of the ideal voltage value respectively H Lower limit U L For voltage values ​​that do not meet the upper and lower limit constraints under special circumstances, a secondary upper limit U′ is set respectively. H Second lower limit U′ L The voltage support capability of the micro-energy storage system after voltage adjustment is evaluated as follows:

[0027]

[0028] Where S V For the full score of the ideal voltage, α l α hThese are the discount factors for the suboptimal voltage; the user experience of the voltage state is quantified over time according to the following formula:

[0029]

[0030] α l S T The time dimensions of discount factor and ideal voltage are respectively the full score of the time dimension. T represents the time set under special experience, denoted as a certain period of night.

[0031] The valuation function for remaining battery capacity is as follows:

[0032]

[0033] Where, α e For discount functions, This represents the difference between the average grid voltage and the lower limit of the ideal voltage over a certain period of time in the future. If the voltage is higher than the lower limit of the ideal voltage, the difference is recorded as 0.

[0034] Furthermore, step 3 specifically includes:

[0035] First, based on the grid state, regulation actions, and voltage regulation value assessment model of the micro-energy storage system constructed in steps 1 and 2, a training network is built: taking the grid state as input and the micro-energy storage regulation actions as output, the network model is trained through a value assessment function; the training network consists of a main network and a target network, and both the main network and the target network include an execution network and an evaluation network. The parameters θ of the execution network and the evaluation network of the main network are randomly selected. π , Perform initialization settings and synchronize the corresponding settings to the target network, i.e., θ π ′←θ π , Set the playback memory unit to an empty set, call the gradient iteration algorithm for N rounds of training, and maximize the evaluation value of the control action.

[0036] Furthermore, the step of calling the gradient iteration algorithm for N rounds of training specifically includes: training and optimizing the parameters of the execution network and evaluation network in the main network and the target network based on gradient iteration, first initializing a Gaussian random process. σ t Using the standard deviation, the grid state described in step 1 is used as the input to the training network, and the execution action of the micro-energy storage system is calculated according to the following formula:

[0037]

[0038] π( s t|θ π ) indicates that when the network parameter is θ πAt that time, the power grid state s t The control strategy, based on the obtained execution action, obtains the value function evaluation value in step 2, and based on the grid state monitoring in step 1, obtains the grid state value at the next moment, and combines the quadruple (s) t a t R t s t+1 Stored in playback memory unit Then, a small batch of samples is randomly extracted from the memory units, and the reference value of the target network is calculated according to the following formula:

[0039]

[0040] Based on this, the loss function and network gradient are updated, and the calculation formula is as follows:

[0041]

[0042]

[0043] The parameters of the execution network and the evaluation network in the target network are updated according to the following formula:

[0044] θ π ′←τθ π +(1-τ)θ π ′

[0045]

[0046] Where τ represents the network update coefficient.

[0047] Furthermore, step 4 specifically includes: determining the sampling period of the micro-energy storage system monitoring module and performing discretization preprocessing on the time; monitoring and extracting the grid voltage state in each time slot, and sending it as input to the training network obtained in step 3; the training network outputs the execution actions of the micro-energy storage system according to the strategy, and sets its working mode and target voltage value through the micro-energy storage system device; achieving voltage active power regulation support through discharge in the low voltage state; and charging the micro-energy storage system battery according to the grid voltage state when it is normal or high, so as to improve the voltage support duration of the micro-energy storage system in the distribution area.

[0048] The advantages of this invention are:

[0049] (1) A voltage regulation strategy for a micro energy storage system is proposed to solve the problem of low voltage in rural power grids. By gradually setting the working mode and target voltage of the micro energy storage system, it is possible to respond to the power grid status in real time and effectively adapt to the power grid support requirements under voltage fluctuation conditions.

[0050] (2) A voltage regulation method for micro energy storage system to provide long-term support for low voltage phenomenon in rural power grid is proposed. Based on the monitoring data of the electrical state of the power grid and the operating state of micro energy storage system, the charging and discharging strategies of the battery are optimized by dynamically regulating the active power of micro energy storage system, thereby improving the effective support time for low voltage phenomenon in rural power grid.

[0051] (3) A voltage regulation algorithm for micro energy storage system based on deep deterministic policy gradient learning is proposed. By modeling the grid observation state, micro energy storage system execution actions and micro energy storage system voltage regulation value during the voltage regulation process, and the optimization strategy based on deep deterministic policy gradient learning, the intelligent solution of the voltage regulation execution actions of micro energy storage system is realized. Attached Figure Description

[0052] Figure 1 This is a system block diagram of a micro-energy storage system voltage regulation method that provides long-term support for low voltage phenomena in rural power grids according to the present invention.

[0053] Figure 2 This is a simplified circuit diagram of a micro-energy storage system connected in series with a rural power grid.

[0054] Figure 3 This is a schematic diagram of a deep deterministic policy gradient learning algorithm. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] The training process for voltage regulation of the micro energy storage system of the present invention will be carried out as follows: Figure 1 The block diagram shown illustrates the process. The state monitoring module first monitors the electrical state of the power grid and the micro-energy storage system, extracts data, and expresses it according to the state model. The state data is input to the learning network, and the output is the action of the micro-energy storage system. The power grid state will respond according to the actual load and the action of the micro-energy storage system. The response result will be fed back to the learning network through a value evaluation function for the optimization learning of network parameters. Only the process of inputting state monitoring data into the network to obtain the action is considered as the control process in practical applications of this method. In summary, the main implementation method of this invention includes the following steps:

[0057] Step 1: Monitor the voltage status of the power grid in areas with frequent low voltage, and based on the voltage regulation characteristics of series micro-energy storage, perform process modeling of the power grid status and regulation actions to provide an information expression model for regulation strategies;

[0058] Step 2: Construct a voltage regulation value assessment model for micro energy storage systems with the goal of long-term voltage support capability;

[0059] Step 3: Based on the grid state, regulation actions, and voltage regulation value assessment model of the micro-energy storage system constructed in Step 1 and Step 2, construct a training network, and train the voltage regulation strategy of the micro-energy storage system through deep deterministic strategy gradient optimization to obtain the training model.

[0060] Step 4: Based on the training model in Step 3 and the actual grid distribution area status information, the working mode and target voltage of the series micro energy storage system are gradually set to achieve active power regulation of the micro energy storage voltage in the actual operating environment.

[0061] Furthermore, in step 1, based on the voltage regulation characteristics of series micro-energy storage, process modeling of the grid state and regulation actions is performed, specifically including:

[0062] Rural power distribution network area connection lines can be abstracted as Figure 2 As shown, the voltage, current, and power values ​​of the micro-energy storage system's installation nodes are acquired through a sensor monitoring device, denoted as U. O I O P O and Q O Based on the following electrical connection formula:

[0063]

[0064] Q O =I O 2 X u

[0065] P O =I O 2 R u +P ESS

[0066] The system resistance R is calculated. u Reactance X u , where P ESS The active power of the micro-energy storage system is positive when charging and negative when discharging. Based on this, a process model of voltage regulation of the micro-energy storage system is constructed, where the grid state model is... s t This represents the power grid state information at time t, including system voltage. Energy storage battery capacity e t Reactance z t and time characteristics δ t ;

[0067] Considering the regulation rules of the micro-energy storage system, the action model is denoted as: m t The operating modes are categorized into three types: charging, discharging, and standby. The target supporting voltage value for the energy storage system;

[0068] Among them, a voltage regulation function is designed to characterize the parameter relationship between the power grid state and the regulation process, denoted as . When the battery capacity meets the requirements

[0069]

[0070] s t (e t ) indicates with The remaining capacity of the energy storage battery after adjusting to the target; when the battery capacity is insufficient to meet the target voltage support requirements, the actual adjustment situation is calculated according to the following formula:

[0071]

[0072] This represents the maximum allowable charging / discharging power of the energy storage at time t.

[0073] Furthermore, the voltage regulation value assessment model for the micro-energy storage system in step 2 separately evaluates the grid voltage support utility and the remaining battery energy, and provides the final evaluation value of the micro-energy storage regulation action in a weighted summation form, providing a basis for network model training for subsequent regulation strategies. Its expression can be written as:

[0074]

[0075] Where ω1 and ω2 are weight parameters, which will be set based on experience, r u (·), r e (·) represent the value of the grid voltage after regulation by the micro-energy storage system and the future supporting value of the remaining battery capacity, respectively; where r u (·)~s v (u), r u (·)~s t (u);

[0076] Set the upper limit U of the ideal voltage value respectively H Lower limit U L For voltage values ​​that do not meet the upper and lower limit constraints under special circumstances, a secondary upper limit U′ is set respectively. H Second lower limit U′L The voltage support capability of the micro-energy storage system after voltage adjustment is evaluated as follows:

[0077]

[0078] Where S V For the full score of the ideal voltage, α l α h These are the discount factors for the suboptimal voltage; the user experience of the voltage state is quantified over time according to the following formula:

[0079]

[0080] α l S T The time dimensions of discount factor and ideal voltage are respectively the full score of the time dimension. T represents the time set under special experience, denoted as a certain period of night.

[0081] The valuation function for remaining battery capacity is as follows:

[0082]

[0083] Where, α e For discount functions, This represents the difference between the average grid voltage and the lower limit of the ideal voltage over a certain period of time in the future. If the voltage is higher than the lower limit of the ideal voltage, the difference is recorded as 0.

[0084] Furthermore, step 3 will construct a training network based on the grid state, micro-energy storage regulation actions, and the micro-energy storage system voltage regulation value assessment model established in steps 1 and 2, to train and optimize the voltage regulation strategy of the micro-energy storage system. Specifically:

[0085] Using the grid state as input and the micro-energy storage regulation action as output, the network model is trained through a value evaluation function. For example... Figure 3 As shown, the training network consists of two parts: a main network and a target network. Both the main network and the target network include an execution network and an evaluation network. The parameters θ of the execution network and the evaluation network of the main network are randomly selected. π , Perform initialization settings and synchronize the corresponding settings to the target network, i.e., θ π ′←θ π , Set the playback memory unit to an empty set, call the gradient iteration algorithm for N rounds of training, and maximize the evaluation value of the control action.

[0086] The gradient iteration algorithm trains and optimizes the parameters of the execution and evaluation networks in both the main and target networks. Specifically, it includes initializing a Gaussian random process. σ t Using the standard deviation, the grid state described in step 1 is used as the input to the training network, and the execution action of the micro-energy storage system is calculated according to the following formula:

[0087]

[0088] π(s t |θ π ) indicates that when the network parameter is θ π At that time, the power grid state s t The control strategy is to obtain the value function evaluation value in step 2 based on the obtained execution action, and obtain the grid state value at the next moment based on the grid state monitoring in step 1. The quadruple (s) t a t R t s t+1 Stored in playback memory unit Then, a small batch of samples is randomly extracted from the memory units, and the reference value of the target network is calculated according to the following formula:

[0089]

[0090] Based on this, the loss function and network gradient are updated, and the calculation formula is as follows:

[0091]

[0092]

[0093] The parameters of the execution network and the evaluation network in the target network are updated according to the following formula:

[0094] θ π ′←τθ π +(1-τ)θ π ′

[0095]

[0096] Where τ represents the network update coefficient.

[0097] Furthermore, step 4, based on the network parameters obtained from training in step 3, connects the micro-energy storage system to monitor fluctuating grid conditions in actual operation, and calls the trained model to gradually set the operating mode and target voltage of the micro-energy storage system. Specifically, this includes: determining the sampling period of the micro-energy storage system monitoring module and performing discretization preprocessing on the time; monitoring and extracting the grid voltage state within each time slot, which is then fed into the training network obtained in step 3; the training network outputs the execution actions of the micro-energy storage system according to the strategy, and sets its operating mode and target voltage value through the micro-energy storage system device; and achieving voltage active power regulation support through discharge in low-voltage conditions. Simultaneously, depending on whether the grid voltage state is normal or high, the micro-energy storage system battery is charged to extend the voltage support duration of the micro-energy storage system in the distribution area.

[0098] This invention addresses the low voltage phenomenon that occurs in remote rural power grid areas under sudden load increases by utilizing a micro-energy storage system to provide long-term voltage support. It effectively enhances the intelligent level of active power regulation of the micro-energy storage system's batteries and has significant value for improving the voltage quality management capabilities of rural power grids.

[0099] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A micro-energy storage voltage active power regulation method for long-term low-voltage support in rural power grids, characterized in that... Includes the following steps: Step 1: Monitor the voltage status of the power grid in areas with frequent low voltage, and based on the voltage regulation characteristics of series micro-energy storage, perform process modeling of the power grid status and regulation actions to provide an information expression model for regulation strategies; Step 2: Construct a voltage regulation value assessment model for micro energy storage systems with the goal of long-term voltage support capability; Step 3: Based on the grid state, regulation actions, and voltage regulation value assessment model of the micro-energy storage system constructed in Step 1 and Step 2, construct a training network, and train the voltage regulation strategy of the micro-energy storage system through deep deterministic strategy gradient optimization to obtain the training model. Step 4: Based on the training model in Step 3 and the actual grid distribution area status information, the working mode and target voltage of the series micro energy storage system are gradually set to achieve active power regulation of the micro energy storage voltage in the actual operating environment. Step 1 involves modeling the grid state and regulation actions based on the voltage regulation characteristics of series micro-energy storage, specifically including: The voltage, current, and power values ​​of the micro-energy storage system's installation nodes are acquired through sensor monitoring devices and denoted as follows: and Based on the following electrical connection formula: ； ； ； The system resistance was calculated. Reactance ,in, The active power of the micro-energy storage system is positive when charging and negative when discharging. Based on this, a process model of voltage regulation of the micro-energy storage system is constructed, where the grid state model is... , express Real-time grid status information, including system voltage Energy storage battery capacity Reactance and time characteristics ; Considering the regulation rules of the micro-energy storage system, the action model is denoted as: , The operating modes are categorized into three types: charging, discharging, and standby. The target supporting voltage value for the energy storage system; Among them, a voltage regulation function is designed to characterize the parameter relationship between the power grid state and the regulation process, denoted as . When the battery capacity meets the requirements, ； Indicates The remaining capacity of the energy storage battery after adjusting to the target; when the battery capacity is insufficient to meet the target voltage support requirements, the actual adjustment situation is calculated according to the following formula: ； express The maximum allowable charge / discharge power for energy storage at any given time; The voltage regulation value assessment model for the micro-energy storage system in step 2 evaluates the grid voltage utility and battery remaining energy separately, and provides the final evaluation value of the micro-energy storage regulation action in a weighted summation form. This provides a basis for network model training for subsequent regulation strategies, and its expression is denoted as: ； in, The weighting parameters are set based on experience. These respectively represent the value of the grid voltage after regulation by the micro-energy storage system and the future supporting value of the remaining battery capacity; among which , ; Set the upper limit of the ideal voltage value respectively Lower limit For voltage values ​​that do not meet the upper and lower limit constraints under special circumstances, a secondary upper limit is set respectively. Second lower limit The voltage support capability of the micro-energy storage system after voltage adjustment is evaluated as follows: ； in This is the full score for the ideal voltage. These are the discount factors for the suboptimal voltage; the user experience of the voltage state is quantified over time according to the following formula: ； The discount factor and the ideal voltage were given full marks in the time dimension, respectively. A set of times representing a special experience, denoted as a certain period of nighttime; The valuation function for remaining battery capacity is as follows: ； in, For discount functions, This represents the difference between the average grid voltage and the lower limit of the ideal voltage over a certain period of time in the future. If the voltage is higher than the lower limit of the ideal voltage, the difference is recorded as 0.

2. The micro-energy storage voltage active power regulation method for long-term low-voltage support of rural power grids as described in claim 1, characterized in that: Step 3 specifically includes: First, based on the grid state, regulation actions, and voltage regulation value assessment model of the micro-energy storage system constructed in steps 1 and 2, a training network is built: taking the grid state as input and the micro-energy storage regulation actions as output, the network model is trained through a value assessment function; the training network consists of a main network and a target network, and both the main network and the target network include an execution network and an evaluation network. The parameters of the execution network and the evaluation network of the main network are randomly selected. Perform initialization settings and synchronize the corresponding settings to the target network, i.e. The playback memory unit is set to an empty set, and the gradient iteration algorithm is called to perform N rounds of training to maximize the evaluation value of the control action.

3. The micro-energy storage voltage active power regulation method for long-term low-voltage support of rural power grids as described in claim 2, characterized in that: The step of calling the gradient iteration algorithm for N rounds of training specifically includes: training and optimizing the parameters of the execution network and evaluation network in the main network and the target network according to the gradient iteration, first initializing a Gaussian random process. , Using the standard deviation, the grid state described in step 1 is used as the input to the training network, and the execution action of the micro-energy storage system is calculated according to the following formula: ； This indicates that the network parameters are At that time, the power grid status The control strategy, based on the obtained execution action, obtains the value function evaluation value in step 2, and based on the monitored grid state in step 1, obtains the grid state value at the next moment, and sets the quadruple... Store in playback memory unit Then, a small batch of samples is randomly extracted from the memory units, and the reference value of the target network is calculated according to the following formula: ； Based on this, the loss function and network gradient are updated, and the calculation formula is as follows: ； ； The parameters of the execution network and the evaluation network in the target network are updated according to the following formula: ； ； in This represents the network update coefficient.

4. The micro-energy storage voltage active power regulation method for long-term low-voltage support of rural power grids as described in claim 1, characterized in that: Step 4 specifically includes: determining the sampling period of the micro-energy storage system monitoring module and performing discretization preprocessing on the time; monitoring and extracting the grid voltage state in each time slot, and feeding it as input into the training network obtained in step 3; the training network outputs the execution actions of the micro-energy storage system according to the strategy, and sets its working mode and target voltage value through the micro-energy storage system device; achieving voltage active power regulation support through discharge in the low voltage state; and charging the micro-energy storage system battery according to the grid voltage state when it is normal or high, so as to improve the voltage support duration of the micro-energy storage system in the distribution area.

Images