Phase selection control method for high-capacity generator circuit breaker

Abstract

The invention discloses a phase selection control method for a high-capacity generator circuit breaker, and belongs to the technical field of high-capacity generator protection equipment. Generator bus current is collected in real time through a Rogowski coil; the identification speed is improved by adopting graded synchronous sampling; based on a current attenuation model containing a steady-state component and a direct-current offset component, short-circuit current is calculated, and a zero crossing point of the short-circuit current is predicted; when a zero crossing point is approached, the operating mechanism and the repulsive force mechanism are controlled to act cooperatively, so that rapid and accurate opening is realized. Through the Newton-Raphson iteration method, the model parameters are adaptively corrected under the condition that the Hessian matrix is approximately updated by adopting the BFGS quasi-Newton method, and the zero crossing point prediction precision and the model robustness are remarkably improved. According to the invention, the problems of serious contact ablation and breaking failure caused by long arcing time when a high-capacity generator circuit breaker breaks a great short-circuit current are solved, the breaking success rate of the circuit breaker can be remarkably increased, the electrical life of the circuit breaker can be remarkably prolonged, and the safety of main equipment can be remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of protection equipment for large-capacity generators, and more specifically, to a phase selection control method for circuit breakers of large-capacity generators. Background Technology

[0002] In power systems, large-capacity generators serve as core power generation equipment, with rated currents typically exceeding 10kA, and short-circuit currents reaching over 100kA during short-circuit faults. If short-circuit faults are not promptly and accurately interrupted, the enormous current can cause severe damage to generators, circuit breakers, and other related power equipment, and may even trigger major safety accidents such as grid paralysis.

[0003] High-capacity generator circuit breakers play a crucial role in protection and control, and their tripping performance directly affects the safe and stable operation of the generator and the entire system. Due to the large time constant of the generator (typically τ > 100ms), there is a significant non-periodic DC offset component in the short-circuit current, causing the current waveform to deviate severely from zero. The first or even the first two cycles do not pass through zero, posing a serious challenge to the accurate tripping of the circuit breaker.

[0004] Traditional circuit breaker tripping control methods are mostly based on fixed delays or simple current amplitude judgments, adopting a "trip immediately after fault detection" mode, without fully considering the phase characteristics of short-circuit current and the attenuation process of DC component. This results in highly random tripping timing, often breaking near the current peak, triggering huge arc energy, causing contact erosion, difficulty in dielectric recovery, easy arc reignition, tripping failure, and even equipment damage or system collapse.

[0005] To overcome the limitations of traditional methods, phase-selective tripping technology has emerged. Its core principle is to achieve precise tripping by predicting the current zero-crossing point, thereby reducing arc energy and improving tripping reliability. However, existing phase-selective control methods are mostly designed for transmission line circuit breakers and do not fully consider special operating conditions such as the complex attenuation characteristics of short-circuit currents in large-capacity generators and significant DC offset, resulting in large errors in zero-crossing point prediction (typically exceeding 10ms). Furthermore, ordinary current transformers are prone to saturation under high current, making it difficult to accurately reflect the initial current characteristics of a fault; traditional sampling methods struggle to balance wide range and high accuracy; and conventional operating mechanisms typically have an action time exceeding 30ms with significant time dispersion, failing to meet the requirements for rapid interruption.

[0006] Therefore, there is an urgent need to develop a phase-selective tripping control method that can adapt to the short-circuit characteristics of large-capacity generators. Summary of the Invention

[0007] To address the problem that existing phase selection and tripping technologies fail to consider the complex short-circuit current attenuation characteristics and significant DC offset of large-capacity generators, leading to large errors in zero-crossing point prediction, this invention proposes a phase selection control method for circuit breakers of large-capacity generators. This method must be able to accurately calculate the current zero-crossing point considering DC attenuation and achieve coordinated action between the operating mechanism and the repulsion mechanism, thereby improving the reliability, speed, and lifespan of the circuit breaker and ensuring the safe and stable operation of the power system.

[0008] According to one aspect of the present invention, a phase selection control method for a high-capacity generator circuit breaker is provided, the circuit breaker comprising a Rogowski coil for sensing generator bus current, an operating mechanism for providing power for opening and closing the circuit breaker, and a repulsion mechanism for providing opening repulsion force, the method comprising the following steps: receiving a generator bus current signal sensed by the Rogowski coil, wherein the sampling frequency of the Rogowski coil is not less than 10 kHz; based on the system rated current... The current signal is sampled synchronously in different ranges, and a short-circuit fault is determined based on the sampling results. When a short-circuit fault occurs, the nearest zero-crossing point of the short-circuit current is calculated based on a preset short-circuit current attenuation model, wherein the short-circuit current attenuation model includes a steady-state AC component and a DC offset component. Before the current zero-crossing point, the operating mechanism and the repulsion mechanism of the circuit breaker are coordinated to achieve rapid clearance of the short-circuit fault within 20ms.

[0009] Based on the above, and more specifically, the short-circuit current attenuation model is constructed as follows: When a short circuit occurs, the short-circuit current decay model is as follows:

[0010] in, This refers to the system voltage amplitude. The equivalent impedance magnitude of the system. Where R is the system equivalent series resistance and L is the system equivalent inductance; The system angular frequency; It is the initial phase angle of the power supply voltage when a short circuit occurs; It is the impedance angle. At this point, the short-circuit current consists of a steady-state AC component and a DC offset component, among which, The formula for calculating steady-state AC components is:

[0011] The formula for calculating the DC offset component is:

[0012] in, The decay time constant, ; It is the load current instant before the short circuit.

[0013] Based on the above, more specifically, numerical iteration or analytical methods are used to solve the equations. ,Sure The smallest positive real root under the given conditions is the most recent time when the current crosses zero. .

[0014] Based on the above, as a further solution, the short-circuit current decay model is adaptively corrected using the Newton-Raphson iteration method, wherein the Hessian matrix required in the Newton-Raphson iteration method is updated approximately using the BFGS quasi-Newton method.

[0015] Based on the above, as a more specific approach, the specific process of adaptively correcting the parameters of the short-circuit current attenuation model using the Newton-Raphson iterative method includes: 1) Based on the short-circuit current decay model, the parameters are... and As parameters to be optimized, a short-circuit current model is constructed:

[0016] 2) The objective function is constructed with the goal of minimizing the sum of squared errors between the sampled current values ​​and the model predictions:

[0017] in A series of current sampling values ​​are collected within a specific time window, for example, the specific time window is 5-10 ms after the short circuit fault is determined; n is the number of sampling points within the time window, n≥10; For the k-th sampling time, k=1,2,...,n; Based on the current and Calculated predicted current value; 3) By finding the objective function right and The partial derivatives are used to obtain the gradient vector. Update parameters along the negative gradient direction and : ,in,

[0018]

[0019] 4) Employ the BFGS quasi-Newton method, using a positive definite symmetric matrix. To approximate the Hessian matrix This ensures that in each iteration... ,in, The parameters for the k-th iteration and The increment, ; The gradient of the k-th iteration The increment, BFGS The update formula is as follows:

[0020] in, I is the identity matrix. 5) Solve the system of linear equations · = , obtain parameter increment = , 6) Update parameters: ; ; The iteration terminates and the optimal parameters are output when any of the following conditions are met. and : ≤0.5° and ≤1ms; objective function ≤ ; When the number of iterations reaches 5.

[0021] Based on the above, more specifically, the segmented synchronous sampling of the current signal includes: using the generator's rated current... Based on this, the current is divided into four levels: No-load gear: 0-20% ; Normal load range: 20%-100% ; Overload setting: 100%-200% ; Fault profile: >200% ; The sampling channels corresponding to each gear level synchronously collect current signals in time, and set corresponding automatic switching thresholds for different gear levels. When the current exceeds the threshold set for the current gear level, it automatically switches to a higher gear level. When the sampled current value exceeds the preset short-circuit judgment threshold and the duration is greater than the preset time threshold, a short-circuit fault is determined to have occurred. The preset time threshold is no greater than 2ms.

[0022] According to a second aspect of the present invention, a high-capacity generator circuit breaker is provided for implementing the phase selection control method for a high-capacity generator circuit breaker described above. The method includes a Rogowski coil, an operating mechanism, a repulsion mechanism, and a control unit. The Rogowski coil is electrically connected to the control unit and is used to transmit a sensed generator bus current signal to the control unit. The operating mechanism and the repulsion mechanism are respectively electrically connected to the control unit and are used to receive action commands sent by the control unit. The control unit is configured to perform segmented synchronous sampling of the received current signal, calculate the nearest current zero-crossing time based on a short-circuit current attenuation model, and issue a coordinated tripping command before the current zero-crossing time.

[0023] The beneficial effects of this application are as follows: 1. This invention classifies short-circuit current for fault judgment and predicts the current zero-crossing point through accurate short-circuit current attenuation modeling, ultimately achieving rapid circuit breaker tripping, reducing damage to equipment from short-circuit faults, significantly improving the circuit breaker's breaking success rate and lifespan, and ensuring the safety of generators and main power grid equipment.

[0024] 2. This invention constructs an accurate short-circuit current attenuation model that includes steady-state AC components and DC offset components, thereby more realistically simulating the transient characteristics of actual short-circuit currents. This overcomes the prediction errors caused by traditional methods that only consider steady-state components, and improves the reliability of fault diagnosis and tripping control.

[0025] 3. This invention uses numerical iteration or analytical methods to solve the current zero-crossing time of the model equation, and can quickly and accurately locate the nearest current zero-crossing time within milliseconds.

[0026] 4. This invention introduces a parameter adaptive correction mechanism, which dynamically adjusts key model parameters through real-time residual comparison, further improving the accuracy of the tripping moment and enhancing the anti-interference and adaptability of the control method.

[0027] 5. This invention adopts a multi-level synchronous sampling and threshold judgment mechanism based on rated current. On the one hand, it improves the sensitivity and accuracy of short-circuit fault detection and avoids misjudgment of normal transient processes. On the other hand, through differentiated threshold settings for different levels, it realizes intelligent identification and rapid response to faults of different degrees and optimizes the protection strategy. Attached Figure Description

[0028] Figure 1 This is a flowchart of the phase selection control method for a large-capacity generator circuit breaker according to the present invention.

[0029] Figure 2 This is a flowchart of the adaptive parameter correction of the short-circuit current decay model using the Newton-Raphson iterative method.

[0030] Figure 3 This is a structural block diagram of the high-capacity generator circuit breaker of the present invention. Detailed Implementation

[0031] like Figure 1 As shown, this invention provides a phase selection control method for a large-capacity generator circuit breaker. The circuit breaker includes a Rogowski coil for sensing the generator bus current, an operating mechanism for providing power to the circuit breaker's opening and closing, and a repulsion mechanism for providing opening repulsion force. The method includes the following steps: receiving the generator bus current signal sensed by the Rogowski coil, wherein the sampling frequency of the Rogowski coil is not less than 10kHz; based on the system rated current... The current signal is sampled synchronously in different ranges, and a short-circuit fault is determined based on the sampling results. When a short-circuit fault occurs, the nearest zero-crossing point of the short-circuit current is calculated based on a preset short-circuit current attenuation model, wherein the short-circuit current attenuation model includes a steady-state AC component and a DC offset component. Before the current zero-crossing point, the operating mechanism and the repulsion mechanism of the circuit breaker are coordinated to achieve rapid clearance of the short-circuit fault within 20ms.

[0032] According to the present invention, when a generator experiences a short-circuit fault, the nearest zero-crossing point of the short-circuit current is calculated based on a preset short-circuit current attenuation model. The construction process of the short-circuit current attenuation model is described below.

[0033] When a generator experiences a short-circuit fault, the circuit can be equivalently represented as a series circuit containing a voltage source (RL). Here, R is the system's equivalent resistance; L is the system's equivalent inductance. It should be noted that the system in this article refers to the power system containing the large-capacity generator, specifically the overall power network consisting of the generator, busbars, connected transmission lines, transformers, and related electrical equipment. The voltage source is the generator's sinusoidal electromotive force, expressed as:

[0034] in: Voltage amplitude; This is the initial phase angle of the power supply voltage when a short circuit occurs.

[0035] According to Kirchhoff's Voltage Law (KVL), the voltage balance equation of the circuit is:

[0036] This equation is a first-order linear non-homogeneous ordinary differential equation, describing the short-circuit current. The pattern of change over time.

[0037] The solution to this differential equation consists of two parts: steady-state components. Its corresponding periodic current after a short circuit; transient component This corresponds to the non-periodic DC offset component after a short circuit.

[0038] Right now, =

[0039] (1) Steady-state components

[0040] The steady-state component is a particular solution to the equation, and it takes the form of a sine function:

[0041] in: The equivalent impedance magnitude of the system. ; It is the impedance angle. .

[0042] (2) Transient components

[0043] The transient component is a homogeneous solution to the equation, in the form of an exponentially decaying function. Its derivation is as follows: In the instant before the short circuit occurs, the circuit is in a steady state, and the current is , The load current before the short circuit, during no-load short circuit. ≈0; Since the inductor current cannot change abruptly, ; Substituting t = 0 into the total current expression:

[0044] Right now:

[0045] The initial values ​​of the transient components are obtained as follows:

[0046] The decay law of transient components is determined by the time constant. Therefore, its complete expression is:

[0047] Adding the steady-state component to the transient component, we obtain the complete expression for the total short-circuit current:

[0048] According to the present invention, the short-circuit current attenuation model is adaptively corrected using the Newton-Raphson iteration method, wherein the Hessian matrix required in the Newton-Raphson iteration method is approximately updated using the BFGS quasi-Newton method. The specific process is as follows: 1) Based on the short-circuit current decay model, the parameters are... and As parameters to be optimized, a short-circuit current model is constructed, and the model is transformed into parameters. and The mapping relationship with the predicted current value lays the foundation for subsequent error calculation and parameter updates:

[0049] 2) The objective function is constructed with the goal of minimizing the sum of squared errors between the sampled current values ​​and the model predictions:

[0050] in This represents a series of current sampling values ​​collected within a specific time window; n is the number of sampling points within the time window, n≥10; For the k-th sampling time, k=1,2,...,n; Based on the current and Calculated predicted current value; 3) By finding the objective function right and The partial derivatives are used to obtain the gradient vector. Update parameters along the negative gradient direction and : ,in,

[0051]

[0052] right The partial derivatives reflect The effect of change on error The partial derivatives reflect The impact of changes on error is assessed; the sign of the partial derivative determines the direction of parameter adjustment, and the magnitude of its absolute value determines the adjustment priority. By summing the error contributions of all sampling points, the reliability of the update direction is ensured. 4) For example Figure 2 As shown, the Newton-Raphson iterative method relies on the Hessian matrix (second-order partial derivative matrix) to solve for the parameter increments, but directly calculating the Hessian matrix is ​​highly complex. Therefore, the BFGS quasi-Newton method is adopted, using a positive definite symmetric matrix. To approximate the Hessian matrix This ensures that in each iteration... ,in, The parameters for the k-th iteration and The increment, ; The gradient of the k-th iteration The increment, BFGS The update formula is as follows:

[0053] in, Let I be the identity matrix. By employing the BFGS quasi-Newton method, the computational complexity is reduced while ensuring the iterative convergence speed. This adapts to the real-time requirements of rapid circuit breaker tripping and avoids missing zero-crossing points due to excessive computation time. 5) Solve the system of linear equations · = , obtain parameter increment = , 6) Update parameters: ; ; The iteration terminates and the optimal parameters are output when any of the following conditions are met. and : ≤0.5° and ≤1ms indicates that the parameter increment is sufficiently small; objective function ≤ This indicates that the error is sufficiently small; when the number of iterations reaches 5, infinite iteration is avoided to ensure real-time performance. Specifically, the preset time step of the iteration is no more than 1ms, and the preset residual threshold is 0.05. The calculation error for the current zero-crossing point does not exceed 2ms.

[0054] In one embodiment of the present invention, a phase selection control method for a large-capacity generator circuit breaker is provided, the method being executed by the control unit of the circuit breaker system.

[0055] First, the control unit receives the generator bus current signal from the Rogowski coil. The sampling frequency of the Rogowski coil is set to 10kHz to ensure that rapid changes in current can be captured.

[0056] Next, the control unit uses the generator's rated current I e Based on this, the corrected current signal is sampled synchronously in different ranges. For example, the current is divided into the following four ranges: No-load gear: 0-20% This corresponds to the current range during the generator startup or shutdown phase. Normal load range: 20%-100% This corresponds to the current range during normal generator operation. Overload setting: 100%-200% This corresponds to the current range for short-term overload operation of the generator; Fault profile: >200% This corresponds to the current range where a short circuit fault may occur.

[0057] The sampling channels for each gear level synchronously acquire current signals over time, and different automatic switching thresholds are set for different gear levels: for example, the threshold for the no-load gear is set to 20%. The normal load threshold is set to 100%. The overload threshold is set to 200%. When the current exceeds the set threshold for the current range, it automatically switches to a higher range to ensure measurement accuracy. When the sampled current value exceeds the preset short-circuit judgment threshold and the duration is greater than the preset time threshold, a short-circuit fault is determined to have occurred. The preset time threshold is less than 2ms.

[0058] Specifically, the preset short-circuit detection threshold should be slightly higher than the overload limit, for example, 210%. According to IEEE C37.013, the generator outlet short-circuit current is typically five times or more than the rated current. However, in phase selection control, a judgment needs to be made at the initial stage of the short circuit (when the current is just rising). Therefore, the threshold should not be too high, otherwise it will delay the tripping timing. Therefore, 210%... It represents a compromise between sensitivity and reliability. Furthermore, even if the current briefly spikes to the short-circuit detection threshold due to, for example, a lightning strike, it will not be considered a short circuit as long as the duration is less than 2ms.

[0059] Once a short circuit is detected, the control unit immediately initiates the short-circuit current attenuation model to calculate the current zero-crossing point. The short-circuit current attenuation model is as follows: .

[0060] The control unit first calls the pre-stored system parameters. , , , Initial calculations are performed. To improve accuracy, an adaptive parameter correction stage is then initiated: using the Newton-Raphson iterative method, with a step size of, for example, 1 ms, the residuals between the model's predicted values ​​and the actual sampled values ​​are compared, and the initial phase angle α and time constant τ are dynamically adjusted until the residual is less than 0.05 times the rated current. Subsequently, the corrected model is solved using numerical iteration or analytical methods. The smallest positive real root is the predicted zero-crossing time t0, and the calculation error is controlled within 2ms.

[0061] Finally, considering the inherent operating time of the circuit breaker's operating mechanism and repulsion mechanism, before the predicted zero-crossing time t0, the control unit simultaneously issues a tripping command to both the operating mechanism and the repulsion mechanism. The operating mechanism and the repulsion mechanism work together to ensure that the moving contact separates near the current zero-crossing point, thereby achieving rapid and successful tripping with low arc energy within 20ms after the fault occurs.

[0062] like Figure 3 As shown, in another embodiment of the present invention, a high-capacity generator circuit breaker is provided to implement the method described in the foregoing embodiments. The circuit breaker includes a Rogowski coil, a control unit, an operating mechanism, and a repulsion mechanism. The Rogowski coil is electrically connected to the control unit and is used to transmit the sensed generator bus current signal to the control unit. The Rogowski coil is responsible for sensing the bus current with high precision and wide bandwidth. The control unit is the core processing module, typically implemented by a high-performance DSP or FPGA, and is responsible for executing all signal processing, fault diagnosis, model calculation, and control logic. The operating mechanism and the repulsion mechanism, as execution units, receive instructions from the control unit and respectively provide the power for opening and closing and additional opening acceleration force to ensure rapid operation.

[0063] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A phase selection control method for a high-capacity generator circuit breaker, the circuit breaker comprising a Rogowski coil for sensing generator bus current, an operating mechanism for providing power to the circuit breaker's opening and closing, and a repulsion mechanism for providing opening repulsion force, characterized in that, Includes the following steps: Receive the generator bus current signal sensed by the Rogowski coil, wherein the sampling frequency of the Rogowski coil is not less than 10kHz; Based on generator rated current The current signal is sampled synchronously in stages, and a short-circuit fault is determined based on the sampling results. When a short-circuit fault occurs, the nearest zero-crossing point of the short-circuit current is calculated based on a preset short-circuit current decay model, wherein the short-circuit current decay model includes a steady-state AC component and a DC offset component. Before the current crosses zero, the operating mechanism and repulsion mechanism of the circuit breaker are coordinated to achieve rapid tripping of the circuit breaker.

2. The phase selection control method for a large-capacity generator circuit breaker according to claim 1, characterized in that, The short-circuit current attenuation model is as follows: When a short circuit occurs, the short-circuit current decay model is as follows: in, This refers to the system voltage amplitude. The equivalent impedance magnitude of the system. Where R is the system equivalent series resistance and L is the system equivalent inductance; The system angular frequency; It is the initial phase angle of the power supply voltage when a short circuit occurs; It is the impedance angle. At this point, the short-circuit current consists of a steady-state AC component and a DC offset component, among which, The formula for calculating steady-state AC components is: The formula for calculating the DC offset component is: in, The decay time constant, ; It is the load current instant before the short circuit.

3. The phase selection control method for a large-capacity generator circuit breaker according to claim 2, characterized in that, Solving equations using numerical iteration or analytical methods ,Sure The smallest positive real root under the given conditions is the most recent time when the current crosses zero. .

4. The phase selection control method for a large-capacity generator circuit breaker according to claim 2, characterized in that, Further includes: The short-circuit current decay model is adaptively corrected using the Newton-Raphson iteration method, and the Hessian matrix used in the Newton-Raphson iteration method is updated approximately using the BFGS quasi-Newton method.

5. The phase selection control method for a large-capacity generator circuit breaker according to claim 4, characterized in that, The specific process of adaptively correcting the parameters of the short-circuit current decay model using the Newton-Raphson iterative method includes: 1) Based on the short-circuit current decay model, the parameters are... and As parameters to be optimized, a short-circuit current model is constructed: 2) The objective function is constructed with the goal of minimizing the sum of squared errors between the sampled current values ​​and the model predictions: in This represents a series of current sample values ​​collected within a specific time window; n is the number of sampling points within the time window, n≥10; For the k-th sampling time, k=1,2,...,n; Based on the current and Calculated predicted current value; 3) By finding the objective function right and The partial derivatives are used to obtain the gradient vector. Update parameters along the negative gradient direction and : ,in, 4) Employ the BFGS quasi-Newton method, using a positive definite symmetric matrix. To approximate the Hessian matrix This ensures that in each iteration... ,in, The parameters for the k-th iteration and The increment, ; The gradient of the k-th iteration The increment, BFGS The update formula is as follows: in, I is the identity matrix. 5) Solve the system of linear equations · = , obtain parameter increment = , 6) Update parameters: ; ; The iteration terminates and the optimal parameters are output when any of the following conditions are met. and : ≤0.5° and ≤1ms; objective function ≤ ; When the number of iterations reaches 5.

6. The phase selection control method for a large-capacity generator circuit breaker according to any one of claims 1 to 5, characterized in that, The synchronous sampling of the current signal in different ranges includes: using the generator's rated current. Based on this, the current is divided into four levels: No-load gear: 0-20% ; Normal load range: 20%-100% ; Overload setting: 100%-200% ; Fault profile: >200% ; The sampling channels corresponding to each gear level synchronously collect current signals in time, and set corresponding automatic switching thresholds for different gear levels. When the current exceeds the threshold set for the current gear level, it automatically switches to a higher gear level. When the sampled current value exceeds the preset short-circuit judgment threshold and the duration is greater than the preset time threshold, a short-circuit fault is determined to have occurred. The preset time threshold is no greater than 2ms.

7. A high-capacity generator circuit breaker, used to implement the phase selection control method for a high-capacity generator circuit breaker as described in any one of claims 1 to 6, characterized in that, It includes a Rogowski coil, an operating mechanism, a repulsion mechanism, and a control unit, among which, The Rogowski coil is electrically connected to the control unit and is used to transmit the sensed generator bus current signal to the control unit. The operating mechanism and the repulsive force mechanism are electrically connected to the control unit, respectively, and are used to receive action commands sent by the control unit; The control unit is configured to perform segmented synchronous sampling of the received current signal, calculate the nearest current zero-crossing time based on the short-circuit current attenuation model, and issue a coordinated tripping command before the current zero-crossing time.

Images