A new energy access distribution network station area three-phase imbalance treatment method and device based on a dynamic phase sensitivity coefficient

Abstract

This invention discloses a method and device for managing three-phase imbalance in distribution network areas with renewable energy access based on dynamic phase sensitivity coefficients, belonging to the field of power system technology. The method collects electrical data from the distribution network area's point of common coupling (PCC) and renewable energy access points, performs defective data identification and removal, and aligns multi-source data. A recursive subspace identification algorithm is used to identify and update the state-space model of the distribution network area online. Based on this model, a dynamic phase sensitivity coefficient is calculated. Then, combining the dynamic phase sensitivity coefficient with model predictive control, an optimization model is established with the goal of minimizing three-phase imbalance, and the optimal control strategy instruction set is solved. The instruction set is sent to the power actuator, where a compensation current is injected into the PCC, and the output state is fed back to the online identification and update process, forming a closed-loop control. This achieves proactive management of three-phase imbalance in the distribution network area. This invention improves the distribution network area's capacity to absorb renewable energy and enhances its safe operation.

Description

Technical Field

[0001] This invention belongs to the field of power system technology, and in particular relates to a method and device for managing three-phase imbalance in distribution network areas for new energy access based on dynamic phase sensitivity coefficient. Background Technology

[0002] With the large-scale integration of new energy sources such as photovoltaic and wind power into the distribution network, their inherent intermittent and random power output characteristics have profoundly altered the power flow distribution of the distribution network areas. In scenarios with high penetration of new energy sources, the net load power of the distribution area exhibits drastic fluctuations in both time and space dimensions. Consequently, the three-phase voltage and current amplitudes at the point of common coupling (PCC) frequently undergo asymmetrical changes, ultimately leading to severe three-phase imbalance in the distribution network area. This three-phase imbalance problem triggers a series of cascading issues, including a surge in transformer losses, line overload, voltage deviation exceeding limits, and relay protection malfunctions, severely impacting the distribution network area's ability to absorb new energy and its safe operation level.

[0003] To address the aforementioned issues, existing technologies have proposed three-phase imbalance compensation devices based on instantaneous reactive power theory. The control strategy of these devices relies on electrical data measured by sensors and data acquisition systems. Their basic working principle is as follows: first, the three-phase voltage and current values ​​of the distribution area are acquired in real time through measuring devices; then, the imbalance component of the current system is calculated based on instantaneous reactive power theory; finally, corresponding compensation commands are generated based on the calculation results to drive the compensation device. The entire control process is a reactive, passive compensation method, meaning that the compensation device can only adjust according to the detected imbalance after a three-phase imbalance has already occurred in the system. This control logic fails to fully consider the dynamic characteristics and state evolution process of the distribution network; the generation of its compensation commands depends entirely on the detection results of the current operating state, rather than the prediction of future states.

[0004] As the rate of fluctuation in renewable energy output continues to accelerate, the inherent defects of this passive compensation method become increasingly apparent. When photovoltaic arrays experience a sudden drop in output due to cloud cover or a significant reversal in wind power, the three-phase imbalance in the distribution area will change drastically within a very short time. However, existing compensation devices are limited by the serial processing delays in data sampling, signal processing, and control calculations, making it impossible for them to keep up with the rapid fluctuations in renewable energy power. Detection lag leads to compensation lag, which in turn further exacerbates the imbalance, creating a vicious cycle of insufficient compensation accuracy. The root cause of these problems lies in the lack of online sensing capabilities for the dynamic characteristics of the distribution network. Existing methods cannot quantify the impact of renewable energy power fluctuations on the three-phase imbalance in the distribution area in real time, nor do they possess the ability to predict the future state of the system. Therefore, it is difficult to achieve accurate and rapid management of three-phase imbalance in scenarios with a high proportion of renewable energy integration. To address this technical challenge, there is an urgent need for a three-phase imbalance management scheme that can proactively sense dynamic changes in the system and implement proactive control. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention proposes a method and device for managing three-phase imbalance in new energy distribution network areas based on dynamic phase sensitivity coefficients, thereby resolving the issues present in the prior art.

[0006] Firstly, to achieve the above objectives, this invention provides a method for managing three-phase imbalance in new energy distribution network areas based on dynamic phase sensitivity coefficients, comprising the following steps: Electrical data of common connection points and new energy access points in the distribution network area are collected, and the collected data are processed for bad data identification and removal, as well as multi-source data alignment. Based on the processed data, a recursive subspace identification algorithm is used to identify and update the state space model of the distribution network area online. Calculate the dynamic phase sensitivity coefficient based on the state-space model; Based on the dynamic phase sensitivity coefficient and the state space model, a predictive optimization model is established with the goal of minimizing the three-phase imbalance, and the predictive optimization model is solved to obtain the optimal control strategy instruction set. The optimal control strategy instruction set is sent to the power actuator, which then outputs a compensation current to the common connection point. The output state of the power actuator is fed back to the online identification and update process of the state space model, thus forming a closed-loop control.

[0007] Optionally, the process of collecting electrical data from the common connection points and new energy access points of the distribution network area, and performing defective data identification and removal processing as well as multi-source data alignment processing on the collected data includes: The three-phase voltage amplitude and three-phase current amplitude at the point of common coupling, the active and reactive power input from new energy sources, and the active and reactive power consumed by the load are collected through the SCADA system and advanced measurement system. Each collected electrical data point is subjected to limit verification, sliding window-based mutation verification, and correlation verification based on power conservation and three-phase voltage balance. Data points that fail any of the verifications are marked as bad data and removed from the original sequence. After removing bad data, the data points are aggregated and aligned according to the standard time grid to obtain a standardized data sequence with a unified timestamp.

[0008] Optionally, based on the processed data, the process of online identification and updating of the state-space model of the distribution network area using a recursive subspace identification algorithm includes: The system input vector is formed by the changes in active power of new energy sources, the changes in reactive power of new energy sources, and the changes in three-phase compensation current at the point of common coupling at the current moment, and the system output vector is formed by the changes in the phase angle of three-phase voltage at the point of common coupling. Recursively move the current input and output vectors to the head of the sliding window and remove the oldest data at the tail of the window. Based on the updated sliding window, recursively calculate the orthogonal projection of the future output data matrix onto the past input and output data matrices; Perform truncated singular value decomposition on the recursively updated projection matrix to extract the extended state sequence; Based on the extended state sequence, the system matrix of the state equation and the system matrix of the output equation are estimated by two recursive least squares methods, respectively, to obtain the state space model of the distribution network area.

[0009] Optionally, the process of calculating the dynamic phase sensitivity coefficient based on the state-space model includes: The steady-state gain matrix is ​​calculated based on the system matrix of the state-space model, and the first column is extracted from the steady-state gain matrix as the sensitivity vector of new energy active power. The phase angle imbalance of the three-phase voltage is determined based on the magnitude of the negative sequence component of the three-phase voltage phasor at the point of common coupling. Calculate the partial derivative of the three-phase voltage phase angle unbalance with respect to the three-phase voltage phase angle to obtain the three-phase voltage phase angle gradient vector; The dynamic phase sensitivity coefficient is obtained by multiplying the three-phase voltage phase angle gradient vector with the new energy active power sensitivity vector.

[0010] Optionally, the process of establishing a predictive optimization model with the objective of minimizing the three-phase imbalance based on the dynamic phase sensitivity coefficient and the state-space model, and solving the predictive optimization model to obtain the optimal control strategy instruction set includes: The predicted value of the three-phase voltage phase angle imbalance is used as the controlled variable, and the prediction equation is constructed by the sum of the product of the dynamic phase sensitivity coefficient and the ultra-short-term predicted change of the new energy active power and the product of the steady-state gain matrix and the change of the system input. The objective function is constructed by summing the weighted sum of the squared values ​​of the predicted three-phase voltage phase angle imbalance at each moment in the time domain and the weighted sum of the squared values ​​of the system input changes. Set equipment capacity constraints for power actuators, as well as voltage amplitude constraints and line current constraints at the point of common coupling. The minimum value of the objective function is obtained by solving for the minimum value under the constraints of equipment capacity and system safety, thus obtaining the optimal control strategy instruction set.

[0011] Optionally, the process of issuing the optimal control strategy instruction set to the power actuator, having the power actuator output a compensation current injected into the point of common coupling, and feeding back the output state of the power actuator to the online identification and update process of the state-space model includes: According to the type of the power actuator, the optimal control strategy instruction set is converted into active power setpoint, reactive power setpoint and three-phase current setpoint and sent to the power actuator; The power actuator updates its internal state equation according to the control command, and drives the converter to output a compensation current to inject into the common connection point through pulse width modulation. The total current after compensation is a three-phase symmetrical balanced positive sequence current. The actual output power of the power actuator and the phase angle state of the three-phase voltage at the common connection point are collected at the next moment to form a feedback information set and returned to the online identification and update process of the state space model.

[0012] Secondly, the present invention also provides a three-phase imbalance mitigation device for new energy distribution network access areas based on dynamic phase sensitivity coefficients, used to implement a three-phase imbalance mitigation method for new energy distribution network access areas based on dynamic phase sensitivity coefficients, the device comprising: The data acquisition unit is used to collect electrical data from the common connection points and new energy access points of the distribution network area, and to perform bad data identification and removal processing and multi-source data alignment processing on the collected data. The central processing and control unit is used to identify and update the state space model of the distribution network area online using a recursive subspace identification algorithm based on the data output by the data acquisition unit, calculate the dynamic phase sensitivity coefficient based on the state space model, and establish a predictive optimization model with the goal of minimizing the three-phase imbalance based on the dynamic phase sensitivity coefficient and the state space model, and solve the predictive optimization model to obtain the optimal control strategy instruction set. The power execution unit is used to receive the optimal control strategy instruction set issued by the central processing control unit, output compensation current to inject into the common connection point, and feed back the output status of the power execution unit to the central processing control unit.

[0013] Thirdly, the present invention also provides a computer terminal device, comprising: One or more processors; A memory, coupled to the processor, for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the steps of the above-mentioned method for managing three-phase imbalance in new energy access distribution network areas based on dynamic phase sensitivity coefficients in the first aspect.

[0014] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, it implements the steps of the method for managing three-phase imbalance in distribution network areas for new energy access based on dynamic phase sensitivity coefficient in the first aspect described above.

[0015] Fifthly, the present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the method for managing three-phase imbalance in distribution network areas for new energy access based on dynamic phase sensitivity coefficients as described in the first aspect above.

[0016] Compared with the prior art, the present invention has the following advantages and technical effects: This invention provides a method for managing three-phase imbalance in distribution network areas with renewable energy access based on dynamic phase sensitivity coefficients. It innovatively introduces dynamic phase sensitivity coefficients as the core quantitative indicator of three-phase imbalance, accurately reflecting the impact of renewable energy power fluctuations on the three-phase imbalance of the distribution network area in real time, thus solving the problem that traditional methods rely on static models and cannot adapt to dynamic system changes. By using recursive subspace identification technology to update the state space model of the three-phase imbalance in the distribution network area online, adaptive tracking of the operating characteristics of the distribution network area is achieved, overcoming the accuracy degradation of traditional fixed models when topology changes or load fluctuations occur. By combining dynamic phase sensitivity coefficients with model predictive control depth, a predictive optimization model for the three-phase imbalance of the distribution network area is constructed, significantly improving the timeliness and accuracy of active and reactive power compensation, and avoiding the limitations of traditional lagging control. This invention also designs a corresponding three-phase imbalance management device for distribution network areas, realizing an integrated solution from data acquisition, processing, optimization to execution, effectively improving the adaptability and safe operation level of the distribution network area to high-proportion renewable energy access. Attached Figure Description

[0017] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a flowchart illustrating a method for managing three-phase imbalance in distribution network areas for new energy access based on dynamic phase sensitivity coefficient, according to an embodiment of the present invention. Figure 2 This is a schematic diagram of a three-phase imbalance control device for new energy access distribution network areas based on dynamic phase sensitivity coefficient, according to an embodiment of the present invention. Detailed Implementation

[0018] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0019] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0020] Example 1 like Figure 1 As shown, this embodiment provides a method for managing three-phase imbalance in new energy distribution network areas based on dynamic phase sensitivity coefficients, including: Electrical data of common connection points and new energy access points in the distribution network area are collected, and the collected data are processed for bad data identification and removal, as well as multi-source data alignment. Based on the processed data, a recursive subspace identification algorithm is used to identify and update the state space model of the distribution network area online. Calculate the dynamic phase sensitivity coefficient based on the state-space model; Based on the dynamic phase sensitivity coefficient and the state space model, a predictive optimization model is established with the goal of minimizing the three-phase imbalance, and the predictive optimization model is solved to obtain the optimal control strategy instruction set. The optimal control strategy instruction set is sent to the power actuator, which then outputs a compensation current to the common connection point. The output state of the power actuator is fed back to the online identification and update process of the state space model, thus forming a closed-loop control.

[0021] Furthermore, the process of collecting electrical data from the common connection points and new energy access points of the distribution network area, and performing defective data identification and removal, as well as multi-source data alignment, includes: The three-phase voltage amplitude and three-phase current amplitude at the point of common coupling, the active and reactive power input from new energy sources, and the active and reactive power consumed by the load are collected through the SCADA system and advanced measurement system. Each collected electrical data point is subjected to limit verification, sliding window-based mutation verification, and correlation verification based on power conservation and three-phase voltage balance. Data points that fail any of the verifications are marked as bad data and removed from the original sequence. After removing bad data, the data points are aggregated and aligned according to the standard time grid to obtain a standardized data sequence with a unified timestamp.

[0022] Specifically, the implementation process of this embodiment includes: Step 1: Multi-source operation data acquisition and processing of distribution network areas Data acquisition in the distribution network is primarily accomplished through SCADA systems and Advanced Measurement Systems (AMIs). Firstly, intelligent sensors, control units, and smart meters installed at the distribution network's points of common coupling, branch lines, and renewable energy grid connection points collect electrical data sequences, including three-phase voltage, current amplitude, and active and reactive power consumption by the load. Each timestamp is then recorded. t k The electrical data sequence is represented as follows: ; In the formula: U a , U b and U c These represent the phase voltage amplitudes (unit: V) of phases a, b, and c at point PCC, respectively. I a , I b and I c These represent the phase current amplitudes (unit: A) of phases a, b, and c at point PCC, respectively. P dg and Q dg These represent the active power and reactive power input from new energy sources (units: kW, kVar). P load and Q load These represent the active power and reactive power consumed by the load, respectively (units: KW, kVar). t k =k*△T , k For sampling index, △T The sampling period is (e.g., 2 seconds).

[0023] I. Data Processing: Because the collected data contains bad data, it needs to be processed, mainly including bad data identification and removal, and data alignment.

[0024] The collected data were validated using limit check functions, mutation checks, and correlation (consistency) checks, and based on... X raw ( t k Generate quality indicator vectors Q ( t k Each element of this vector corresponds to the quality status of a data point (e.g., 0 = good, 1 = invalid, 2 = out of bounds, 4 = mutation, 8 = inconsistent).

[0025] 1. Limit Value Validation for X raw ( t k Any single measurement value in ) X i ( t k ), and define its reasonable upper and lower limits as [ X iMin , X iMax To determine whether the data is within the specified range, the corresponding limit check function is as follows: f limit ( X i ( t k ))=1(Bad),if X i ( t k ) [ X iMin , X iMax or X i ( t k ) is null;0(Good),otherwise.

[0026] 2. Mutation verification Define a sliding window of length L, and calculate the relative rate of change of the current value to the previous valid value and the deviation of robust statistics within the window.

[0027] The verification process is as follows: First, calculate the window.W ={ X i ( t k-L ),…, X i ( t k-1 )} M Secondly, calculate the absolute deviation D of all data within the window from the median. X j - M |, X j ∈ W Then, calculate the median of the absolute deviation. MAD In addition, calculate the current value. X i ( t k ) and median M Standardized absolute deviation Where c is a constant, usually taken as 1.4826, such that MAD Standard deviation of normal distribution σ The process is as follows; finally, a mutation check function is used to determine whether a mutation has occurred. The corresponding mutation check function is as follows: f spike ( X i ( t k ))={1(Bad), if Z > ξ ; 0(Good),otherwise}; in: ζ This is the threshold used for mutation detection, typically set to 3-5.

[0028] 3. Relevance (consistency) verification Power conservation verification and three-phase voltage balance verification are performed based on Kirchhoff's laws and the principle of power conservation.

[0029] a) Power conservation check: At point PCC, the difference between the input power of the renewable energy source and the power consumed by the load should be conserved relative to network losses: P net ( t k )= P dg ( t k )- P load ( tk )≈ P loss ( t k ); Q net ( t k )= Q dg ( t k )- Q load ( t k )≈ Q loss ( t k ); In the formula: P net ( t k ) indicates in t k The net active power input to the PCC point at any given time; Q net ( t k ) indicates in t k The net reactive power input to the PCC point at any given time; P loss ( t k ) indicates in t k Active power loss in the distribution network area at all times; Q loss ( t k ) indicates in t k Reactive power loss in the distribution network area at all times.

[0030] Net input power P dg ( t k ) and power P measured ( t k The absolute difference between the two values ​​is defined as the residual of the power conservation check. r p ( t k ): r p ( t k )=|(P dg ( t k )- P load ( t k ))- P measured ( t k )|; In the formula: P measured ( t k It can be calculated from voltage and current, and the corresponding expression is: P measured = U a * I a *cos θ a + U b * I b *cos θ b + U c * I c *cos θ c ; Will r p ( t k Substitute the values ​​into the power conservation check function for verification. The corresponding power conservation check function is as follows: f consistency ( r p ( t k ))={1(BAD),if r p ( t k )> η P ;0(Good),otherwise}; In the formula: η P This indicates the threshold set based on measurement error and line loss estimation.

[0031] b) Three-phase voltage balance verification: Calculate the unbalance of three-phase voltage: ; ; In the formula: U avg ( t k ) indicates in t k The arithmetic mean of the three-phase voltage amplitudes at point PCC; U a ( t k ), U b ( t k )and U c ( t k ) represent PCC points respectively t k The phase voltage amplitudes of phases a, b, and c at time points; U unbalance ( t k ) indicates in t k The three-phase voltage imbalance at a given time.

[0032] The three-phase voltage balance verification function is as follows: f unbalance ( U unbalance ( t k ))={1(Bad),if U unbalance ( t k )> δ U ;0(Good),otherwise}; In the formula: δ U This indicates the voltage imbalance threshold.

[0033] Finally, by combining all the discriminant functions, the final quality label is generated: ; when f limit , f spike , f consistency and f unbalance If any of the tests fails, then the corresponding Q( t k It is marked as an exception (Bad).

[0034] Data points marked as "Bad" are removed from the original sequence to form the sequence to be interpolated. X gap (t k ).

[0035] II. Data Alignment: Because multiple devices use different frequencies and clocks during data acquisition, the data needs to be aligned to a standard time grid.

[0036] For each standard time point From a time window Aggregate and calculate the valid data within.

[0037] for X gap ( t k Each data point in ) X i Its alignment time value Generated by the following aggregate functions: ; in: This is the intermediate sequence containing data gaps after removing bad data; For standard timestamps The top-aligned data vector; Aggregation represents the aggregation function.

[0038] Furthermore, based on the processed data, the process of online identification and updating of the state-space model of the distribution network area using the recursive subspace identification algorithm includes: The system input vector is formed by the changes in active power of new energy sources, the changes in reactive power of new energy sources, and the changes in three-phase compensation current at the point of common coupling at the current moment, and the system output vector is formed by the changes in the phase angle of three-phase voltage at the point of common coupling. Recursively move the current input and output vectors to the head of the sliding window and remove the oldest data at the tail of the window. Based on the updated sliding window, recursively calculate the orthogonal projection of the future output data matrix onto the past input and output data matrices; Perform truncated singular value decomposition on the recursively updated projection matrix to extract the extended state sequence; Based on the extended state sequence, the system matrix of the state equation and the system matrix of the output equation are estimated by two recursive least squares methods, respectively, to obtain the state space model of the distribution network area.

[0039] Specifically, the implementation process of this embodiment includes: Step 2: Online identification and updating of the three-phase imbalance state space model of the distribution network area First, k -1 and k The change in electrical data (active power, reactive power, three-phase current, etc.) at any given time is defined as a column vector of dimension 5. : ; In the formula: △ P dg ( k ) and △ Q dg ( k ) represent the changes in active power and reactive power input from new energy sources, respectively; △I comp,a (k), △I comp,b (k) and △I comp,c (k) represents the input current changes of phases a, b, and c at the point of common coupling (PCC) of the compensation device (such as a smart soft switch SOP), respectively, as the control input to be optimized.

[0040] The control target (the phase angle of the three-phase voltage at the PCC point) is defined as a 3-dimensional column vector. y ( k The corresponding vectors are as follows: y ( k )=[ θ a ( k ), θ b ( k ), θ c ( k )] T ; In the formula: θ a ( k ), θ b ( k )and θ c ( k The phase angles of phases A, B, and C at the point of common coupling (PCC) are respectively relative to the phase angles of phases B and C. k The change in the sampled value at time -1 (unit: radians).

[0041] Secondly, the time-varying dynamic characteristics of the distribution network substations are captured in real time, and the state space model of the distribution network substations is updated online based on the recursive subspace identification algorithm. The specific process is as follows: 1. Recursive Update of Data Vector: The recursive subspace identification algorithm recursively updates a sliding window of length L to cache the latest input data. u ( i ) and output data y ( i When a new set of data u ( k )and y ( k When the data is recursively moved to the head of the sliding window, the oldest data at the tail of the window will be removed, forming a continuously updated data window.

[0042] 2. Recursive Calculation of Projection and Decomposition: Based on the updated data window, a recursive projection method and incremental matrix factorization technique are used to update the future output space. Y f past input / output space Orthographic projection on: ; In the formula, Y f For future output space or future output data matrix; W p This is a matrix of combined input / output data from the past. U p For past input data matrix; Y p Output data matrix for the past; Ŷ f express Y f exist W p Orthogonal projection on.

[0043] 3. Extraction of the extended state sequence: Perform truncated singular value decomposition (SVD) on the recursively updated projection matrix: ; in: This represents a diagonal matrix composed of the principal singular values. U n and V n These represent the left and right singular vector matrices, respectively. Extended state sequence. for: ; 4. Recursive least squares estimation of the system matrix: through extended state sequences Xf The system matrix is ​​estimated based on two recursive least squares (RLS) methods. The RLS problem corresponding to the state equation is as follows: ; in, x ( k ) indicates the physical network of the distribution network area. k The state vector at any given time; u ( k )for k The system input vector at any given time (e.g., power output of the transformer area, power output of new energy sources, etc.); A is the state matrix; B is the input matrix.

[0044] The RLS problem corresponding to the output equation: ; In the formula: y ( k )express k The output vector of the distribution network area at any given time (such as data on the bus voltage, line current, total power supply, etc. of the area). C This is the output matrix; D It is a feedforward matrix.

[0045] 5. Finally, the identified state-space model of the distribution network area {A, B, C, D} is obtained: ; ; In the formula: w ( k )express k Random disturbances in the distribution network area at all times; v ( k )express k Errors and interference introduced by time sensors.

[0046] Furthermore, the process of calculating the dynamic phase sensitivity coefficient based on the state-space model includes: The steady-state gain matrix is ​​calculated based on the system matrix of the state-space model, and the first column is extracted from the steady-state gain matrix as the sensitivity vector of new energy active power. The phase angle imbalance of the three-phase voltage is determined based on the magnitude of the negative sequence component of the three-phase voltage phasor at the point of common coupling. Calculate the partial derivative of the three-phase voltage phase angle unbalance with respect to the three-phase voltage phase angle to obtain the three-phase voltage phase angle gradient vector; The dynamic phase sensitivity coefficient is obtained by multiplying the three-phase voltage phase angle gradient vector with the new energy active power sensitivity vector.

[0047] Specifically, the implementation process of this embodiment includes: Step 3: Online calculation of dynamic phase sensitivity coefficient (DPSC) Based on the distribution network substation state-space model constructed in step 2 { A , B、C、D The specific steps for calculating the dynamic phase sensitivity coefficient (DPSC) are as follows: 1. Calculate the steady-state gain matrix G For the state-space model of a distribution network area { A , B、C、D}, its steady-state gain matrix as follows: G = C ( I - A ) -1 B + D ; In the formula: I Representation and state matrix A An identity matrix of the same dimension.

[0048] 2. Extract the power sensitivity vector of new energy sources Z p From the steady-state gain matrix G Extract the first column to form the new energy active power sensitivity vector. Z p : ; Where: vector Z p The new energy power sensitivity vector represents the effect of a unit change in the active power of new energy on the phase angle of each phase voltage at point PCC.

[0049] 3. Calculate the three-phase voltage phase angle unbalance index. I θ Based on the symmetrical component method, the three-phase voltage phasors can be expressed as follows: V a = e (jθa) , V b = e (jθb) and V c = e (jθc) The corresponding negative order components V 2 for: ; In the formula: O=e (j120°) For rotating operators, the phase angle imbalance of three-phase voltages Iθ This indicates the negative-order component. V Amplitude of 2: I θ =|V 2 | ; In the formula: I θ The larger the value, the more severe the three-phase imbalance in the distribution network area.

[0050] 4. Calculate the phase angle gradient vector of the three-phase voltage. J :Will I θ For three phase angle variables θ a , θ b and θ c Taking the partial derivatives separately, we get: ; In the formula: the phase angle gradient vector of the three-phase voltage J Characterizing the effect of changes in the phase angle of the three-phase voltage in a distribution network area on the three-phase unbalance index I θ The extent of the impact.

[0051] 5. Calculate the Dynamic Phase Sensitivity Coefficient (DPSC): Using the chain rule, obtain the dynamic phase sensitivity coefficient (DPSC) for three-phase imbalance in the distribution network area: .

[0052] Furthermore, based on the dynamic phase sensitivity coefficient and the state-space model, the process of establishing a predictive optimization model with the objective of minimizing the three-phase imbalance, and solving the predictive optimization model to obtain the optimal control strategy instruction set includes: The predicted value of the three-phase voltage phase angle imbalance is used as the controlled variable, and the prediction equation is constructed by the sum of the product of the dynamic phase sensitivity coefficient and the ultra-short-term predicted change of the new energy active power and the product of the steady-state gain matrix and the change of the system input. The objective function is constructed by summing the weighted sum of the squared values ​​of the predicted three-phase voltage phase angle imbalance at each moment in the time domain and the weighted sum of the squared values ​​of the system input changes. Set equipment capacity constraints for power actuators, as well as voltage amplitude constraints and line current constraints at the point of common coupling. The minimum value of the objective function is obtained by solving for the minimum value under the constraints of equipment capacity and system safety, thus obtaining the optimal control strategy instruction set.

[0053] Specifically, the implementation process of this embodiment includes: Step 4: Optimization Solution Based on DPSC and Model Predictive Control Considering the control costs of managing three-phase imbalance in distribution network areas (including equipment wear and tear, energy loss, and operating costs), an optimization model is established with the objective of minimizing the three-phase imbalance in the distribution network areas. The corresponding objective function is: ; In the formula: N p To predict the length of the time domain; N c To control the time domain length and satisfy N c ≤N p ; I θ,pred ( k + i|k ) indicates in k Always k+i The predicted value of the three-phase voltage phase angle imbalance of the distribution network area at a given time; Δ u ( k+i | k )express k Time k+ i The change in the input variables of the state-space model of the distribution network area at any given time; Q It is a 3*3 positive definite weight matrix used to weigh the effect of suppressing the three-phase imbalance of the distribution network area; It is a 5x5 positive definite weight matrix used to weigh control costs.

[0054] The relevant constraints include equipment capacity constraints and system security constraints: (1) Equipment capacity constraints: ; in: u ( k+i )=[ P comp ( k+i ), Q comp ( k+i ),I comp , a ( k+i ), I comp,b ( k+i ), I comp,c ( k+i )] T Indicates in k+i Input variables for the state-space model of the distribution network area at any given time; P comp ( k+i )and Q comp ( k+i ) respectively represent in k + i The active and reactive power output by the constant compensation device.

[0055] (2) System security constraints: ; ; In the formula: U pcc ( k+i ) indicates in k+i Voltage amplitude at point PCC at any given time; U min and U max These are the minimum and maximum allowable voltage values, determined according to the power system operation regulations; I line ( k+i ) indicates in k+i Line current at any given time; I line,max This indicates the maximum allowable current carrying capacity of the line, which can be determined by the line's thermal stability limit.

[0056] By minJ( k The instruction set for the three-phase imbalance control strategy of the distribution network area was obtained and substituted into the three-phase imbalance prediction model of the distribution network area for verification. The prediction model is as follows: I θ,pred ( k + i | k )= I θ ( k )+ DPSC ( k )*Δ P dg,pred ( k + i )+ J * G *Δu( k + i|k ); In the formula: I θ ( k )express k The phase angle imbalance of the three-phase voltage in the distribution network area at any given time; DPSC ( k )for k The dynamic phase sensitivity coefficient at time; △ Pdg,pred ( k+i () represents the ultra-short-term predicted change in the active power of new energy sources; J G is a 1*3 three-phase voltage phase angle gradient vector; G is a 3*5 steady-state gain matrix.

[0057] After verification, the optimal control strategy instruction set for the three-phase imbalance management of the distribution network area can be obtained, including current setpoint, active power setpoint and reactive power setpoint.

[0058] When the power actuator is a smart soft switch (SOP), the optimal control strategy instruction set for three-phase imbalance mitigation in the distribution network area is as follows: ; In the formula: P ref ( k ) indicates the active power setting value of SOP (in kW). Q ref ( k This indicates the reactive power setting value of SOP (in Kvar). I com,a,ref ( k ), I com,b,ref ( k )and I com,c,ref ( k ) represent the three-phase current settings of phases a, b, and c at the point of common coupling (PCC) of SOP, respectively.

[0059] When the power actuator is a static synchronous compensator (STATCOM), the optimal control strategy instruction set for three-phase imbalance mitigation in the distribution network area is as follows: .

[0060] Furthermore, the process of issuing the optimal control strategy instruction set to the power actuator, having the power actuator output a compensation current injected into the point of common coupling, and feeding back the output state of the power actuator to the online identification and update process of the state-space model includes: According to the type of the power actuator, the optimal control strategy instruction set is converted into active power setpoint, reactive power setpoint and three-phase current setpoint and sent to the power actuator; The power actuator updates its internal state equation according to the control command, and drives the converter to output a compensation current to inject into the common connection point through pulse width modulation. The total current after compensation is a three-phase symmetrical balanced positive sequence current. The actual output power of the power actuator and the phase angle state of the three-phase voltage at the common connection point are collected at the next moment to form a feedback information set and returned to the online identification and update process of the state space model.

[0061] Specifically, the implementation process of this embodiment includes: Step 5: Issuance and closed-loop execution of the optimal control strategy instruction set Based on step 3, the optimal control strategy instruction set for three-phase imbalance mitigation in each distribution network area is obtained and issued to the corresponding power actuator. Upon receiving the optimal control strategy instruction set, the power actuator... U 1 ( k After that, the driving state equation x c ( k Updated as follows: x c ( k )= A c* x c ( k -1)+ B c* U 1 ( k ); in: x c ( k -1) indicates that the power actuator is in k The state vector at time -1 has specific state variables that depend on the equipment structure and typically include: AC side inductor current of the converter, DC bus capacitor voltage, phase-locked loop output phase, and intermediate states of the internal controller, etc. A c This represents the state matrix of the power actuator; B c Indicates external control command U 1 Impact on power actuators.

[0062] Actual output power P act ( k The state equation of the power actuator x c ( k ) and the dynamic characteristic matrix of the power actuator itself C c Decide: P act ( k )= Cc* x c ( k ); in: C c Indicates the power actuator's output power P act The impact; P act ( k ) indicates that the power actuator is in k The power vector that is actually input into the distribution network at any given time.

[0063] Based on the intelligent soft-switching power actuator, the three-phase compensation current commands considering active and reactive compensation currents are as follows: Active current compensation command i p,ref,abc ( k ) and reactive current compensation command i q,ref,abc ( k Based on the power actuator of the static synchronous compensator, a reactive current compensation command considering the reactive current compensation is obtained. i q,ref,abc ( k ).

[0064] After receiving a current compensation command, the pulse width modulation (PWM) unit and power switching devices of the intelligent soft-switching and static synchronous compensator power actuator drive the converter through closed-loop control (such as hysteresis control or proportional resonant control) to make its AC side output a compensation current consistent with the command. i comp,abc ( k ).

[0065] According to Kirchhoff's current law, the compensation current... i comp,abc ( k After injecting the PCC point, the following conditions must be met: i grid,abc ( k )= i load,abc ( k )+ i dg,abc ( k )+ i comp,abc ( k ); In the formula: i load,abc ( k () represents the three-phase current of the load in the distribution network area. i dg,abc (k This represents the three-phase current injected into the distribution network area by new energy sources. i grid,abc ( k The number () represents the three-phase current of the distribution network area after compensation by the power actuator, i.e., the total current after compensation. i grid,abc ( k ( ) is a three-phase symmetrical, balanced positive sequence current, and the three phases of the distribution network area are balanced.

[0066] Finally, the output power at time k+1 is based on the intelligent soft-switching and static synchronous compensator power actuator. P act ( k+ 1) Phase angle status of three-phase voltage at PCC point in the distribution network y ( k+ 1) Establish a feedback information set including changes in power, current, and voltage: F ( k+ 1)={ P act ( k+ 1), y ( k+ 1)}; The information set is returned to step 2 for a new round of rolling optimization to achieve closed-loop control of the three-phase imbalance management of the distribution network area.

[0067] Example 2 like Figure 2 As shown, based on the same general inventive concept, this invention also provides a three-phase imbalance mitigation device for new energy distribution network access areas based on dynamic phase sensitivity coefficients. The following describes the three-phase imbalance mitigation device for new energy distribution network access areas based on dynamic phase sensitivity coefficients provided by this invention. The three-phase imbalance mitigation device for new energy distribution network access areas based on dynamic phase sensitivity coefficients described below can be referred to in correspondence with the three-phase imbalance mitigation method for new energy distribution network access areas based on dynamic phase sensitivity coefficients described above. The device includes: The data acquisition unit is used to collect electrical data from the common connection points and new energy access points of the distribution network area, and to perform bad data identification and removal processing and multi-source data alignment processing on the collected data. The central processing and control unit is used to identify and update the state space model of the distribution network area online using a recursive subspace identification algorithm based on the data output by the data acquisition unit, calculate the dynamic phase sensitivity coefficient based on the state space model, and establish a predictive optimization model with the goal of minimizing the three-phase imbalance based on the dynamic phase sensitivity coefficient and the state space model, and solve the predictive optimization model to obtain the optimal control strategy instruction set. The power execution unit is used to receive the optimal control strategy instruction set issued by the central processing control unit, output compensation current to inject into the common connection point, and feed back the output status of the power execution unit to the central processing control unit.

[0068] Example 3 (equivalent to Example 2 above) In this embodiment, a computer terminal device is provided, including: One or more processors; A memory, coupled to the processor, for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the steps of the above-described method for managing three-phase imbalance in distribution network areas for new energy access based on dynamic phase sensitivity coefficients.

[0069] In this embodiment, a computer-readable storage medium is also provided, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of the above-described method for managing three-phase imbalance in new energy access distribution network areas based on dynamic phase sensitivity coefficients.

[0070] In this embodiment, a computer program product is also provided, including a computer program that, when executed by a processor, implements the steps of the above-described method for managing three-phase imbalance in new energy access distribution network areas based on dynamic phase sensitivity coefficients.

[0071] This invention provides a method for managing three-phase imbalance in distribution network areas with renewable energy access based on dynamic phase sensitivity coefficients. It innovatively introduces dynamic phase sensitivity coefficients as the core quantitative indicator of three-phase imbalance, accurately reflecting the impact of renewable energy power fluctuations on the three-phase imbalance of the distribution network area in real time, thus solving the problem that traditional methods rely on static models and cannot adapt to dynamic system changes. By using recursive subspace identification technology to update the state space model of the three-phase imbalance in the distribution network area online, adaptive tracking of the operating characteristics of the distribution network area is achieved, overcoming the accuracy degradation of traditional fixed models when topology changes or load fluctuations occur. By combining dynamic phase sensitivity coefficients with model predictive control depth, a predictive optimization model for the three-phase imbalance of the distribution network area is constructed, significantly improving the timeliness and accuracy of active and reactive power compensation, and avoiding the limitations of traditional lagging control. This invention also designs a corresponding three-phase imbalance management device for distribution network areas, realizing an integrated solution from data acquisition, processing, optimization to execution, effectively improving the adaptability and safe operation level of the distribution network area to high-proportion renewable energy access.

[0072] The above are merely preferred embodiments 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 scope of the technology 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 method for managing three-phase imbalance in distribution network areas for new energy access based on dynamic phase sensitivity coefficient, characterized in that, Includes the following steps: Electrical data of common connection points and new energy access points in the distribution network area are collected, and the collected data are processed for bad data identification and removal, as well as multi-source data alignment. Based on the processed data, a recursive subspace identification algorithm is used to identify and update the state space model of the distribution network area online. Calculate the dynamic phase sensitivity coefficient based on the state-space model; Based on the dynamic phase sensitivity coefficient and the state space model, a predictive optimization model is established with the goal of minimizing the three-phase imbalance, and the predictive optimization model is solved to obtain the optimal control strategy instruction set. The optimal control strategy instruction set is sent to the power actuator, which then outputs a compensation current to the common connection point. The output state of the power actuator is fed back to the online identification and update process of the state space model, thus forming a closed-loop control.

2. The method according to claim 1, characterized in that, The process of collecting electrical data from common connection points and new energy access points in the distribution network area, and then identifying and removing defective data and aligning multi-source data includes: The three-phase voltage amplitude and three-phase current amplitude at the point of common coupling, the active and reactive power input from new energy sources, and the active and reactive power consumed by the load are collected through the SCADA system and advanced measurement system. Each collected electrical data point is subjected to limit verification, sliding window-based mutation verification, and correlation verification based on power conservation and three-phase voltage balance. Data points that fail any of the verifications are marked as bad data and removed from the original sequence. After removing bad data, the data points are aggregated and aligned according to the standard time grid to obtain a standardized data sequence with a unified timestamp.

3. The method according to claim 2, characterized in that, Based on the processed data, the process of online identification and updating of the state-space model of the distribution network area using the recursive subspace identification algorithm includes: The system input vector is formed by the changes in active power of new energy sources, the changes in reactive power of new energy sources, and the changes in three-phase compensation current at the point of common coupling at the current moment, and the system output vector is formed by the changes in the phase angle of three-phase voltage at the point of common coupling. Recursively move the current input and output vectors to the head of the sliding window and remove the oldest data at the tail of the window. Based on the updated sliding window, recursively calculate the orthogonal projection of the future output data matrix onto the past input and output data matrices; Perform truncated singular value decomposition on the recursively updated projection matrix to extract the extended state sequence; Based on the extended state sequence, the system matrix of the state equation and the system matrix of the output equation are estimated by two recursive least squares methods, respectively, to obtain the state space model of the distribution network area.

4. The method according to claim 3, characterized in that, The process of calculating the dynamic phase sensitivity coefficient based on the state-space model includes: The steady-state gain matrix is ​​calculated based on the system matrix of the state-space model, and the first column is extracted from the steady-state gain matrix as the sensitivity vector of new energy active power. The phase angle imbalance of the three-phase voltage is determined based on the magnitude of the negative sequence component of the three-phase voltage phasor at the point of common coupling. Calculate the partial derivative of the three-phase voltage phase angle unbalance with respect to the three-phase voltage phase angle to obtain the three-phase voltage phase angle gradient vector; The dynamic phase sensitivity coefficient is obtained by multiplying the three-phase voltage phase angle gradient vector with the new energy active power sensitivity vector.

5. The method according to claim 4, characterized in that, Based on the dynamic phase sensitivity coefficient and the state-space model, the process of establishing a predictive optimization model with the objective of minimizing the three-phase imbalance, and solving the predictive optimization model to obtain the optimal control strategy instruction set includes: The predicted value of the three-phase voltage phase angle imbalance is used as the controlled variable, and the prediction equation is constructed by the sum of the product of the dynamic phase sensitivity coefficient and the ultra-short-term predicted change of the new energy active power and the product of the steady-state gain matrix and the change of the system input. The objective function is constructed by summing the weighted sum of the squares of the predicted values ​​of the three-phase voltage phase angle imbalance at each time point in the time domain and the weighted sum of the squares of the changes in the system input. Set equipment capacity constraints for power actuators, as well as voltage amplitude constraints and line current constraints at the point of common coupling. The minimum value of the objective function is obtained by solving for the minimum value under the constraints of equipment capacity and system safety, thus obtaining the optimal control strategy instruction set.

6. The method according to claim 5, characterized in that, The process of issuing the optimal control strategy instruction set to the power actuator, having the power actuator output a compensation current injected into the point of common coupling, and feeding back the output state of the power actuator to the online identification and update process of the state space model includes: According to the type of the power actuator, the optimal control strategy instruction set is converted into active power setpoint, reactive power setpoint and three-phase current setpoint and sent to the power actuator; The power actuator updates its internal state equation according to the control command, and drives the converter to output a compensation current to inject into the common connection point through pulse width modulation. The total current after compensation is a three-phase symmetrical balanced positive sequence current. The actual output power of the power actuator and the phase angle state of the three-phase voltage at the common connection point are collected at the next moment to form a feedback information set and returned to the online identification and update process of the state space model.

7. A three-phase imbalance mitigation device for new energy access distribution network areas based on dynamic phase sensitivity coefficient, characterized in that, The apparatus for implementing the method according to any one of claims 1-6 comprises: The data acquisition unit is used to collect electrical data from the common connection points and new energy access points of the distribution network area, and to perform bad data identification and removal processing and multi-source data alignment processing on the collected data. The central processing and control unit is used to identify and update the state space model of the distribution network area online using a recursive subspace identification algorithm based on the data output by the data acquisition unit, calculate the dynamic phase sensitivity coefficient based on the state space model, and establish a predictive optimization model with the goal of minimizing the three-phase imbalance based on the dynamic phase sensitivity coefficient and the state space model, and solve the predictive optimization model to obtain the optimal control strategy instruction set. The power execution unit is used to receive the optimal control strategy instruction set issued by the central processing control unit, output compensation current to inject into the common connection point, and feed back the output status of the power execution unit to the central processing control unit.

8. A computer terminal device, characterized in that, include: One or more processors; A memory, coupled to the processor, for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors perform the steps of the method as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1-6.

Images