Transformer parallel cooperative voltage regulation method and system based on multi-objective dynamic optimization
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- STATE GRID HUNAN ELECTRIC COMPANY DISASTER PREVENTION & REDUCTION CENT
- Filing Date
- 2026-05-22
- Publication Date
- 2026-08-07
AI Technical Summary
该方式存在固有缺陷：第一，调节为有级调压，响应速度慢，无法实现精细、连续调节；第二，机械动作机构存在磨损和故障风险，维护成本高；第三，更关键的是其控制策略通常较为简单，往往仅以某一点的电压为控制目标，或进行简单的均流控制，难以实现“抑制环流”、“均衡负载率”和“降低系统损耗”等多目标的实时协同最优
[0054] (1) Significantly improved regulation performance: By utilizing power electronics technology, millisecond-level response and stepless continuous voltage regulation are achieved, eliminating the inherent disadvantages of mechanical voltage regulation.
Smart Images
Figure CN122533004A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of power system operation and control technology, and in particular to a transformer parallel coordinated voltage regulation method and system based on multi-objective dynamic optimization. Background Technology
[0002] In substations, data centers, and large industrial plants and mines, multiple transformers are often connected in parallel to improve power supply capacity and reliability. Transformers operating in parallel must meet conditions such as equal turns ratios, identical connection groups, and similar per-unit short-circuit voltages. However, in actual operation, slight differences in equipment parameters, tap positions, or uneven line impedance can lead to circulating currents between the transformers. These circulating currents are primarily reactive, do not increase load power supply capacity, but increase transformer winding losses, exacerbate equipment heating, and in severe cases, even affect power supply safety.
[0003] The traditional method for regulating the voltage of parallel transformers relies on the transformer's own on-load tap changer. This method has inherent drawbacks: First, the regulation is stepped, resulting in a slow response speed and an inability to achieve precise and continuous regulation; second, the mechanical operating mechanism is susceptible to wear and failure, leading to high maintenance costs; third, and more importantly, its control strategy is usually relatively simple, often focusing only on the voltage at a single point or performing simple current sharing control, making it difficult to achieve real-time optimal coordination among multiple objectives such as "suppressing circulating current," "balancing load rate," and "reducing system losses."
[0004] In recent years, the application of power electronics technology in voltage regulation has become increasingly widespread. For example, there has been research on using series compensators (such as dynamic voltage restorers, DVRs) for stepless voltage regulation of single transformers or lines. However, existing technologies mostly focus on the voltage compensation function of individual devices, such as solving voltage sags or performing single-point voltage stabilization. When applied to the specific scenario of parallel transformer banks, simply connecting multiple independent voltage regulating modules in parallel lacks a global coordinated control strategy. Each module adjusts only based on local information, which may have limited effectiveness in suppressing circulating currents and may even cause new oscillation problems due to conflicting control objectives, failing to achieve overall optimal operation at the system level. Summary of the Invention
[0005] To suppress circulating current, optimize load distribution, and reduce system losses, this application provides a transformer parallel coordinated voltage regulation method and system based on multi-objective dynamic optimization.
[0006] Firstly, this application provides a transformer parallel coordinated voltage regulation method based on multi-objective dynamic optimization, employing the following technical solution:
[0007] A method for coordinated voltage regulation of transformer parallel connections based on multi-objective dynamic optimization includes the following steps:
[0008] S1: Real-time data of electrical quantities and bus voltage of each parallel transformer branch are obtained through the state sensing unit.
[0009] S2: Obtain short-term load forecast information and combine it with the real-time data from step S1 to construct a multi-objective optimization model with the core objectives of suppressing circulating current and optimizing load distribution.
[0010] S3: Solve the multi-objective optimization model to obtain the target voltage command values on the primary side of each parallel transformer that make the objective function optimal;
[0011] S4: The power electronic voltage regulating module generates a corresponding compensation voltage, which is applied to the primary side of the corresponding transformer to precisely control its secondary side output voltage.
[0012] Optionally, the state variables and measurements required to construct the multi-objective optimization model include:
[0013] Power grid bus voltage ;
[0014] Primary and secondary currents of each transformer , Primary and secondary voltages and load current wait;
[0015] The calculated load rates of each transformer Active power loss wait;
[0016] Construct the objective function J of the collaborative control center:
[0017]
[0018] in, The set weighting coefficients, For circulation suppression target, For load balancing purposes; The target for system loss.
[0019] Optional, circulating current suppression target Jcir: minimize reactive circulating current among transformers; this can be achieved by calculating the primary current and average current of each transformer. Characterized by the sum of squares of the deviations:
[0020]
[0021] The load balancing objective Jload is to ensure that each transformer bears the load proportionally according to its capacity, achieving optimal allocation.
[0022] ;
[0023] in This is the ideal load factor for the i-th transformer. Its rated capacity;
[0024] System loss target Minimize the total system losses, including transformer copper and iron losses and switching and conduction losses of the voltage regulating module. Transformer losses can be calculated using its equivalent circuit model and real-time current, while voltage regulating module losses can be estimated based on device characteristics and modulation models.
[0025]
[0026] in Indicates the copper and iron losses of the transformer. This indicates the switching and conduction losses of the voltage regulation module.
[0027] Optionally, the objective function J is constrained as follows:
[0028] Equality constraints: Kirchhoff's voltage and current laws are determined by the topology of the circuit network.
[0029] Inequality constraints include the allowable range of grid voltage, the safe upper and lower limits of transformer primary voltage, the allowable range of load rate for each transformer, and the compensation voltage and current of each voltage regulating module within the rated values of its power devices, etc.
[0030] Safe range of primary voltage of transformer:
[0031]
[0032] Voltage regulation module output capacity limitations:
[0033] Voltage regulator module current limit:
[0034] Maximum load factor for each transformer:
[0035] Limits on the rate of change of control variables in adjacent periods: .
[0036] Optionally, a load prediction sequence Iload(k+1), ..., Iload(k+NP) for the next NP control cycles can be provided, which allows the optimization model to be upgraded from single-point optimization to finite-time rolling optimization.
[0037] At each time k, the control center solves an optimal control problem based on the current and predicted future states, and the objective function becomes:
[0038] ;
[0039] After optimization, only the optimal control command u(k) at the current time k is issued, and this process is repeated in the next cycle.
[0040] Ultimately, the solution algorithm outputs the optimal control variable sequence u(k) that minimizes the objective function J, where the first element Vcomp,i(k) is the precise voltage command issued to the i-th voltage regulation module at the current moment.
[0041] Secondly, this application also provides a transformer parallel coordinated voltage regulation system based on multi-objective dynamic optimization, which adopts the following technical solution:
[0042] A transformer parallel coordinated voltage regulation system based on multi-objective dynamic optimization includes:
[0043] Collaborative Control Center;
[0044] The status sensing unit is used to collect voltage, current, and power information of each branch of the parallel transformer group;
[0045] The power electronic voltage regulating module includes a rectifier, a DC capacitor, an inverter, a filter, a controller, and a series coupling transformer; it adopts a series compensation topology, and its output is connected in series with the primary winding of the corresponding parallel transformer and the grid connection point through the coupling transformer;
[0046] The collaborative control center is communicatively connected to the state sensing unit and each of the power electronic voltage regulating modules. It executes a multi-objective optimization model algorithm based on state sensing data and load forecast information, and issues voltage regulation commands to each power electronic voltage regulating module.
[0047] Optionally, the collaborative control center has a built-in load prediction module, which is used to generate a short-term load prediction curve based on historical operating data and external environment information, and use the prediction information as the feedforward input of the multi-objective optimization model.
[0048] Optionally, the controller in the power electronic voltage regulating module includes:
[0049] Coordinate transformation unit: used to transform between the stationary coordinate system and the synchronous rotating coordinate system, and its phase reference is determined by the synchronous phase angle of the grid voltage provided by the phase-locked loop circuit;
[0050] Voltage outer loop controller: It takes external displacement voltage command and local feedback voltage as input, and its output serves as the reference command for inner loop current.
[0051] Inner current loop controller: It takes the current reference command output from the outer voltage loop and the local feedback current as inputs. Its output is decoupled and fed forward to form the modulation voltage command of the inverter bridge.
[0052] Pulse Width Modulation Unit: Converts modulation voltage commands into switching signals to drive fully controllable power devices in the module.
[0053] In summary, this application includes the following beneficial technical effects:
[0054] (1) Significantly improved regulation performance: By utilizing power electronics technology, millisecond-level response and stepless continuous voltage regulation are achieved, eliminating the inherent disadvantages of mechanical voltage regulation.
[0055] (2) Significantly enhanced operational safety: Through active collaborative control with the core objective of suppressing reactive circulating current, the circulating current can be reduced or even eliminated from the source, reducing equipment thermal stress and improving system stability.
[0056] (3) Optimization of economic operation level: Taking load balancing and minimum loss as clear optimization objectives, the parallel transformer group always operates at or near the highest efficiency point, so as to achieve energy saving and consumption reduction.
[0057] (4) Possesses forward-looking adjustment capability: It innovatively introduces short-term load forecasting, enabling optimization decisions not only to be based on the current state but also to predict future trends, conduct "forward-looking" adjustments, smooth the operating trajectory, avoid frequent or oscillating adjustments, and improve the level of intelligence in control. Attached Figure Description
[0058] Figure 1 This is a schematic diagram of a voltage intelligent control system for multiple transformers operating in parallel, as proposed in this invention.
[0059] Figure 2 This is a schematic diagram of a power electronic voltage regulating module proposed in this invention. Detailed Implementation
[0060] The following is in conjunction with the appendix Figure 1-2 This application will be described in further detail.
[0061] This application discloses a transformer parallel coordinated voltage regulation method based on multi-objective dynamic optimization, including the following steps:
[0062] S1: Real-time data of electrical quantities and bus voltage of each parallel transformer branch are obtained through the state sensing unit.
[0063] S2: Obtain short-term load forecast information and combine it with the real-time data from step S1 to construct a multi-objective optimization model with the core objectives of suppressing circulating current and optimizing load distribution.
[0064] S3: Solve the multi-objective optimization model to obtain the target voltage command values on the primary side of each parallel transformer that make the objective function optimal;
[0065] S4: The power electronic voltage regulating module generates a corresponding compensation voltage, which is applied to the primary side of the corresponding transformer to precisely control its secondary side output voltage.
[0066] The decision-making process of the collaborative control center is "perception-prediction-optimization-decision," and its core lies in solving a dynamic multi-objective optimization problem to determine the specific voltage command value to be issued to each voltage regulation module.
[0067] u=[Vcomp,1,Vcomp,2,…,Vcomp,n]T
[0068] Where Vcomp,i=Vcomp,i∠θcomp,i is the compensation voltage vector to be determined for the i-th power electronic voltage regulating module, and n is the number of parallel transformers.
[0069] The state variables and measurements required for the model include:
[0070] Power grid bus voltage ;
[0071] Primary and secondary currents of each transformer , Primary and secondary voltages and load current wait;
[0072] The calculated load rates of each transformer Active power loss etc.; Construct the objective function J of the collaborative control center:
[0073]
[0074] Where w1, w2, and w3 are weight coefficients set according to the system's running priority, and the objective function is the weighted sum of the three sub-objectives, where:
[0075] (1) Circulating current suppression target Jcir: Minimize the reactive circulating current among the transformers. This can be characterized by calculating the sum of squares of the deviations (focusing on the reactive component) between the primary current and the average current of each transformer:
[0076] ;
[0077] (2) Load balancing objective Jload: to enable each transformer to bear the load according to its capacity ratio, so as to achieve optimal allocation.
[0078] ;
[0079] in Let be the ideal load rate of the i-th transformer, and Sn,i be its rated capacity.
[0080] (3) System loss target Minimize the total system losses, including transformer copper and iron losses and switching and conduction losses of the voltage regulating module. Transformer losses can be calculated using its equivalent circuit model and real-time current, while voltage regulating module losses can be estimated based on device characteristics and modulation models.
[0081] ;
[0082] in Indicates the copper and iron losses of the transformer. This indicates the switching and conduction losses of the voltage regulation module.
[0083] The objective function J is constrained as follows:
[0084] (1) Equality constraints: Kirchhoff's voltage and current law equations determined by the topology of the circuit network.
[0085] (2) Inequality constraints, including the allowable range of grid voltage, the safe upper and lower limits of transformer primary voltage, the allowable range of load rate of each transformer, and the compensation voltage and current of each voltage regulating module within the rated values of its power devices, etc.
[0086] Safe range of primary voltage of transformer:
[0087]
[0088] Voltage regulation module output capacity limitations:
[0089] Voltage regulator module current limit:
[0090] Maximum load factor for each transformer:
[0091] Limits on the rate of change of control variables in adjacent periods: .
[0092] The load prediction module in the collaborative control center provides a load prediction sequence Iload(k+1), ..., Iload(k+NP) for the next NP control cycles, which allows the optimization model to be upgraded from single-point optimization to finite-time-domain rolling optimization.
[0093] At each time k, the control center solves an optimal control problem based on the current and predicted future states, and the objective function becomes:
[0094] ;
[0095] After optimization, only the optimal control command u(k) at the current time k is issued, and this process is repeated in the next cycle. This method enables the adjustment command to have "foresight," smooths the adjustment process, and avoids frequent or over-adjustments caused by load inertia.
[0096] Ultimately, the solution algorithm outputs the optimal control variable sequence u(k) that minimizes the objective function J, where the first element Vcomp,i(k) is the precise voltage command issued to the i-th voltage regulation module at the current moment.
[0097] This application also provides a transformer parallel coordinated voltage regulation system for implementing the above method; comprising:
[0098] Collaborative Control Center;
[0099] The status sensing unit is used to collect voltage, current, and power information of each branch of the parallel transformer group;
[0100] The power electronic voltage regulating module includes a rectifier, a DC capacitor, an inverter, a filter, a controller, and a series coupling transformer; it adopts a series compensation topology, and its output is connected in series with the primary winding of the corresponding parallel transformer and the grid connection point through the coupling transformer;
[0101] The collaborative control center is communicatively connected to the state perception unit and each of the power electronic voltage regulation modules. It is used to execute a multi-objective optimization model algorithm based on the state perception data and load prediction information, and to issue voltage regulation commands to each power electronic voltage regulation module.
[0102] The state-aware units are equipped with measuring devices on the primary and secondary sides of T1 and T2, and the data is uploaded to the collaborative control center via a high-speed communication network. The software algorithm flow running within the collaborative control center is as follows:
[0103] (1) Data synchronization and preprocessing (cycle: 10 milliseconds): Receive the synchronous phasor data (voltage, current amplitude and phase) of each measurement point, and calculate the actual load rate, active / reactive power and circulating current value between transformers of each transformer.
[0104] (2) Obtain forecast information: Obtain the load change forecast curve for the next 5 minutes from an independent load forecasting system or internal forecasting algorithm module.
[0105] (3) Constructing an optimization problem (period: 100 milliseconds): Constructing a multi-objective optimization model with the core of suppressing circulating current and optimizing load distribution, solving the optimization model, and obtaining the target voltage command values of the primary side of each parallel transformer that make the objective function optimal;
[0106] (4) Send each target voltage command value to the corresponding power electronic voltage regulating module;
[0107] (5) Each power electronic voltage regulating module generates a corresponding compensation voltage based on the received instruction, and after being combined with the grid voltage in series, it is applied to the primary side of the corresponding transformer, thereby accurately controlling its secondary side output voltage.
[0108] The power electronic voltage regulating module includes a rectifier, DC capacitor, inverter, filter, controller, and series-coupled transformer. The controller includes:
[0109] (1) Coordinate transformation unit: used to transform between the stationary coordinate system and the synchronous rotating coordinate system. Its phase reference is determined by the synchronous phase angle of the grid voltage provided by the phase-locked loop circuit.
[0110] (2) Voltage outer loop controller: It takes the external displacement voltage command and the local feedback voltage as inputs, and its output serves as the reference command for the inner loop current.
[0111] (3) Current inner loop controller: The current reference command output by the voltage outer loop and the local feedback current are used as inputs. After decoupling and feedforward compensation, its output forms the modulation voltage command of the inverter bridge.
[0112] (4) Pulse width modulation unit: converts the modulation voltage command into a switching signal to drive the fully controlled power device in the module.
[0113] The operating steps of the power electronic voltage regulating module include:
[0114] S1: Command Reception and Coordinate Transformation. The local controller receives compensation voltage commands from the collaborative control center. Simultaneously, the phase-locked loop (PLL) monitors the phase angle θ of the grid voltage in real time. Through Park coordinate transformation, the compensation voltage command is converted from the stationary coordinate system to the synchronously rotating dq coordinate system, obtaining the command component in DC form. and Simultaneously, the output current Iinj and output voltage Vcomp of the current and voltage sensor acquisition module are converted into feedback components using the same transformation angle θ. , and , .
[0115] S2: Voltage outer loop regulation generates current reference. The voltage outer loop controller compares command components. , With feedback components , This generates an error signal. After proportional-integral regulation, this error signal is output as a reference command for the inner current loop. and .
[0116] S3: Current inner loop regulation and dynamic compensation. Reference instruction for current inner loop controller comparison. , With feedback components , Generate preliminary modulation voltage commands. , To eliminate coupling between dq axis variables and enhance disturbance rejection capability, the preliminary modulation voltage command is decoupled and feedforward compensated to generate the final modulation voltage command. , The method is as follows:
[0117] ;
[0118] ;
[0119] in , For decoupling terms, and This is the feedforward term for the transformed grid voltage. This stage ensures that the output current follows the reference command quickly and accurately, forming the basis for the system's high dynamic performance.
[0120] S4: Modulation signal generation and power drive. This involves generating the compensated modulation voltage command. , The system is transformed back to the stationary coordinate system via inverse Park transformation, generating a two-phase modulated wave. This modulated wave is then processed by a space vector pulse width modulation algorithm to produce multiple pulse width modulation signals, which are used to drive the insulated-gate bipolar transistor switches in the full-bridge inverter circuit within the module. Finally, these signals are synthesized at the module output to produce a signal consistent with the command. Consistent compensation voltage.
[0121] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.
Claims
1. A method for parallel coordinated voltage regulation of transformers based on multi-objective dynamic optimization, characterized in that: Includes the following steps: S1: Real-time data of electrical quantities and bus voltage of each parallel transformer branch are obtained through the state sensing unit. S2: Obtain short-term load forecast information and combine it with the real-time data from step S1 to construct a multi-objective optimization model with the core objectives of suppressing circulating current and optimizing load distribution. S3: Solve the multi-objective optimization model to obtain the target voltage command values on the primary side of each parallel transformer that make the objective function optimal; S4: The power electronic voltage regulating module generates a corresponding compensation voltage, which is applied to the primary side of the corresponding transformer to precisely control its secondary side output voltage.
2. The transformer parallel coordinated voltage regulation method based on multi-objective dynamic optimization according to claim 1, characterized in that: The state variables and measurements required to construct the multi-objective optimization model include: Power grid bus voltage ; Primary and secondary currents of each transformer , Primary and secondary voltages and load current ; The calculated load rates of each transformer Active power loss ; Construct the objective function J of the collaborative control center: in, The set weighting coefficients, For circulation suppression target, For load balancing purposes; The target for system loss.
3. The transformer parallel coordinated voltage regulation method based on multi-objective dynamic optimization according to claim 2, characterized in that: The circulating current suppression target Jcir is to minimize the reactive circulating current among the transformers; this is achieved by calculating the primary current of each transformer. With average current Characterized by the sum of squares of the deviations: Load balancing goals This ensures that each transformer bears the load according to its capacity ratio, achieving optimal load distribution. ； in This is the ideal load factor for the i-th transformer. Its rated capacity; System loss target Minimize the total system losses, including transformer copper and iron losses and switching and conduction losses of the voltage regulating module. Transformer losses can be calculated using its equivalent circuit model and real-time current, while voltage regulating module losses can be estimated based on device characteristics and modulation models. in Indicates the copper and iron losses of the transformer. This indicates the switching and conduction losses of the voltage regulation module.
4. The transformer parallel coordinated voltage regulation method based on multi-objective dynamic optimization according to claim 3, characterized in that: The objective function J is constrained as follows: Equality constraints: Kirchhoff's voltage and current laws are determined by the topology of the circuit network. Inequality constraints include the allowable range of grid voltage, the safe upper and lower limits of transformer primary voltage, the allowable range of load rate for each transformer, and the compensation voltage and current of each voltage regulating module within the rated values of its power devices, etc. Safe range of primary voltage of transformer: Voltage regulation module output capacity limitations: Voltage regulator module current limit: Maximum load factor for each transformer: Limits on the rate of change of control variables in adjacent periods: .
5. The transformer parallel coordinated voltage regulation method based on multi-objective dynamic optimization according to claim 4, characterized in that: It provides a load prediction sequence Iload(k+1), ... Iload(k+NP) for the next NP control cycles, which allows the optimization model to be upgraded from single-point optimization to finite-time rolling optimization. At each time k, the control center solves an optimal control problem based on the current and predicted future states, and the objective function becomes: ； After optimization, only the optimal control command u(k) at the current time k is issued, and this process is repeated in the next cycle. Ultimately, the solution algorithm outputs the optimal control variable sequence u(k) that minimizes the objective function J, where the first element Vcomp,i(k) is the precise voltage command issued to the i-th voltage regulation module at the current moment.
6. A transformer parallel coordinated voltage regulation system based on multi-objective dynamic optimization, used to implement the transformer parallel coordinated voltage regulation method based on multi-objective dynamic optimization as described in claim 5, characterized in that: include: Collaborative Control Center; The status sensing unit is used to collect voltage, current, and power information of each branch of the parallel transformer group; The power electronic voltage regulating module includes a rectifier, a DC capacitor, an inverter, a filter, a controller, and a series coupling transformer; it adopts a series compensation topology, and its output is connected in series with the primary winding of the corresponding parallel transformer and the grid connection point through the coupling transformer; The collaborative control center is communicatively connected to the state perception unit and each of the power electronic voltage regulation modules. It is used to execute a multi-objective optimization model algorithm based on the state perception data and load prediction information, and to issue voltage regulation commands to each power electronic voltage regulation module.
7. A transformer parallel coordinated voltage regulation system based on multi-objective dynamic optimization according to claim 6, characterized in that: The collaborative control center has a built-in load prediction module, which generates a short-term load prediction curve based on historical operating data and external environment information, and uses the prediction information as the feedforward input of the multi-objective optimization model.
8. A transformer parallel coordinated voltage regulation system based on multi-objective dynamic optimization according to claim 7, characterized in that: The controller in the power electronic voltage regulating module includes: Coordinate transformation unit: used to transform between the stationary coordinate system and the synchronous rotating coordinate system, and its phase reference is determined by the synchronous phase angle of the grid voltage provided by the phase-locked loop circuit; Voltage outer loop controller: It takes external displacement voltage command and local feedback voltage as input, and its output serves as the reference command for inner loop current. Inner current loop controller: It takes the current reference command output from the outer voltage loop and the local feedback current as inputs. Its output is decoupled and fed forward to form the modulation voltage command of the inverter bridge. Pulse Width Modulation Unit: Converts modulation voltage commands into switching signals to drive fully controllable power devices in the module.