Method for diagnosing a small fault of a current sensor of a three-phase three-level rectifier

Abstract

The application provides a current sensor micro-fault diagnosis method of a three-phase three-level rectifier, and belongs to the field of fault diagnosis.The method comprises the following steps: constructing a voltage and current basic relation formula, establishing a current state equation and a micro-fault equation and expressing the current state equation and the micro-fault equation in an augmented form, constructing a new type of non-singular terminal sliding mode observer, giving a fault state threshold value and a detection characteristic quantity and performing fault diagnosis.The method expresses the current state equation and the micro-fault equation of the original system in an augmented form to perform fault reconstruction, so that the influence of the fault on the system is more intuitively displayed.The method uses the new type of non-singular terminal sliding mode observer to reach the sliding mode surface more quickly and reduce the chattering of the sliding mode motion through reasonable parameter design, so that a better tracking effect of the actual situation is achieved; the method uses an adaptive threshold value to replace a traditional constant threshold value, prevents missing reports or false reports, and improves the robustness of fault diagnosis.

Description

Technical Field

[0001] This invention relates to the field of fault diagnosis, and in particular to a method for diagnosing minor faults in the current sensor of a three-phase three-level rectifier. Background Technology

[0002] The traction rectifier in a high-speed rail traction drive system is a core component and is highly susceptible to failure, potentially causing serious personal injury and property damage. The traction rectifier is a three-phase, three-level rectifier. Currently, there are two main methods for diagnosing minor faults in the current sensors of three-phase, three-level rectifiers:

[0003] 1. Circuit Model-Based Diagnostic Methods. These methods typically involve two stages: residual generation and diagnostic decision-making. First, a hybrid logic dynamic model of the inverter is established. An estimated value of the faulty system output is obtained through an observer. The difference between the estimated value and the true value is used to obtain the residual. Then, the fault category is determined based on corresponding decision rules. Relevant papers and patents include "Research on Fault Diagnosis and Fault-Tolerant Control Based on Sliding Mode Technology and Its Application in High-Speed ​​Trains" (Zhang Kangkang, Nanjing University of Aeronautics and Astronautics, December 2018) and the Chinese invention patent application "A Method for Diagnosing Minor Faults in the Current Sensor of a Static Converter for New Energy Electric Vehicles" (CN 113534035 A). These methods have clear model mechanisms, are easy to implement, and can perform real-time diagnosis. However, the challenge lies in establishing an accurate analytical mathematical model of the control system and maximizing the model's reliability.

[0004] 2. Signal Feature-Based Diagnostic Methods. Due to the difficulty in determining which diagnostic model is suitable for historical fault data or statistical datasets caused by different signals, and the challenge of establishing mathematical models for complex components or systems in researching many practical fault prediction problems, historical sensor data from various stages of system design, simulation, operation, and maintenance become the primary means of judging system performance—a data-driven diagnostic method. This process is divided into training and testing phases. Relevant papers and patents include "Data-Driven Incipient Fault Detection via Canonical Variate Dissimilarity and Mixed Kernel Principal Component Analysis" (Wu Ping, IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, August 2021) and Chinese invention patent application publication "A Design Method for a Micro-Fault Diagnosis System for High-Speed ​​Railway Inverters" (CN 106959397A). These methods do not require a precise mathematical model of the system, but they do require a large amount of accurate system data and appropriate data diagnostic methods.

[0005] In summary, existing technologies suffer from problems such as long diagnostic cycles, large computational load, slow approach rates and significant jitter during sliding mode motion, poor anti-interference ability of adaptive thresholds, high requirements for mathematical models, and poor robustness. Summary of the Invention

[0006] The technical problem to be solved by this invention is the problem mentioned in the background art. Specifically, it proposes a method for diagnosing minor faults in the current sensor of a three-phase three-level rectifier. First, the state equation and fault of the original system are represented in an augmented form, and the fault is reconstructed to make the impact of the fault on the system more intuitive. Second, a novel non-singular terminal sliding mode observer is used with reasonable parameter design to reach the sliding surface faster and reduce chattering of the sliding mode movement, thereby achieving a better tracking effect of the actual situation. At the same time, when diagnosing faults based on the characteristic quantity of current, an adaptive threshold is designed to suppress uncertain disturbances inherent to the system or externally, thereby increasing the accuracy of the diagnostic method and improving the diagnostic precision.

[0007] To achieve the above objectives, the present invention provides a method for diagnosing minor faults in the current sensor of a three-phase three-level rectifier. The circuit topology involved in the method includes a three-phase grid-side voltage source, a three-phase grid-side equivalent inductance and a three-phase grid-side equivalent resistance, a three-phase three-level rectifier, two identical supporting capacitors, a DC-side load, a current sensor, and a control module. The two supporting capacitors are connected in series and then in parallel between the positive DC bus and the negative DC bus of the DC-side load.

[0008] The three-phase three-level rectifier is divided into three phase arms, each connected in parallel with the DC-side load. The three phase arms are denoted as k-phase arms, where k represents the phase sequence (k = a, b, c). Each phase arm consists of four switching transistors connected in series with anti-parallel diodes. Therefore, the three-phase three-level rectifier contains a total of 12 switching transistors with anti-parallel diodes. These 12 switching transistors are denoted as Vswitching-transistor. kn , n represents the serial number of the switching transistor, n = 1, 2, 3, 4; in each phase arm of the three-phase bridge, the switching transistor V k1 Switching transistor V k2 Switching transistor V k3 Switching transistor V k4 In series, the switching transistor V k2 and switching transistor V k3 The connection point is denoted as the rectifier bridge input point τ. k ;

[0009] Let the equivalent inductance of the three-phase grid side be denoted as L. k The equivalent resistance of the three-phase grid side is denoted as R. k The three-phase grid voltage source is denoted as E. k The equivalent inductance L on the three-phase grid side k One end is connected to the rectifier bridge input point τ k The other end is connected to the equivalent resistance R of the three-phase grid. k The equivalent resistance R of the three-phase grid side k The other end is connected to the three-phase grid voltage source E k The positive terminal, the three-phase grid-side voltage source E k The negative terminal is grounded;

[0010] The current sensor's detection terminals are divided into three phases, denoted as detection terminals Γ. k Detection end Γ k Connected to the rectifier bridge input point τ k Equivalent inductance L of the three-phase grid side k Between them, the output terminal of the current sensor is connected to the input terminal of the control module, and the output terminal of the control module is respectively connected to 12 switching transistors V. kn ;

[0011] The diagnostic method includes the following steps:

[0012] Step 1: Denote the three-phase three-level rectifier as a rectifier, and establish the hybrid logic dynamic model of the rectifier, the expression of which is: in, The phase voltage U of the k-phase bridge arm ko The estimated value, δ k Let be the switching function of the k-phase bridge arm;

[0013] Step 2, sample the equivalent resistance R flowing through the three-phase grid side. k The three-phase currents are denoted as the network-measured equivalent current i. a i b i c E at the sampling three-phase grid side voltage source k The three-phase voltage, denoted as grid-side voltage U. a U b U c Sample the voltage at the DC-side load and record it as the DC-side voltage U. dc ;

[0014] The state-space equations of the rectifier are established, and their expression is as follows:

[0015]

[0016] in, For the equivalent current i of the network measurement a i b i c The derivative of R, where R is the resistance value of the equivalent resistance Rk on the three-phase grid side, and L is the equivalent inductance L on the three-phase grid side. k The inductance value, D is the coefficient matrix 1, F represents the unknown disturbance signal of the rectifier;

[0017] Step 3, define the minor fault of the current sensor as minor fault f, and establish the minor fault equation, the expression of which is: in, A is the derivative of the minor fault f. f Let ξ be the Hurwitz matrix, and ξ be the excitation signal for a minor fault.

[0018] Step 4: Use the state augmentation method to establish an augmented system for the small fault equations and the rectifier's state-space equations. The expression of the augmented system is as follows:

[0019]

[0020] Step 5, given the state variable x i i = 1, 2, 3, 4, state variable x i The expression is:

[0021]

[0022] The state variable x i The derivative is denoted as the state variable derivative. Then the derivative of the state variable The expressions for the augmented system output y are as follows;

[0023]

[0024]

[0025] Step 6, set the state variable x i The estimated value is denoted as the state variable estimate. State variable estimates The derivative is denoted as the derivative of the state variable estimate. The estimated value of the augmented system output y is denoted as the output estimate. Construct a novel nonsingular terminal sliding mode observer, whose expression is:

[0026]

[0027]

[0028] Where g is an adjustable parameter of 1, and g > 0; k(t) is the adaptive law, and the expression for the derivative of k(t) is: t is a time variable, representing the running time of the driving system; a is the adaptive law gain coefficient. λ is an adjustable parameter 2, σ is an adjustable parameter 3, both λ and σ are positive odd numbers and satisfy 1 < λ / σ < 2, β is an adjustable parameter 4, β ∈ (0, 1); s is a non-singular terminal sliding surface. r i For residuals, Let be the derivative of the residual, and tanh() be the hyperbolic tangent function.

[0029] Step 7, define the residual r i , Then the residual r i derivative The expression is:

[0030]

[0031] Step 8, given the current sensor minor fault state threshold T th1 Current sensor fault state threshold T th2 and current sensor failure threshold T th3 ;

[0032] Step 9, Define the fault detection feature quantity F sIts expression is:

[0033] F s =(sign(||r4||-T) th1 )+1) / 2+(sign(||r4||-T th2 )+1) / 2+(sign(||r4||-T th3 )+1) / 2

[0034] Based on the fault detection characteristic quantity F s Diagnosing current sensor faults:

[0035] If F s =0, then the current sensor is considered to be in normal condition;

[0036] If F s =1, then the current sensor is considered to be in a minor fault state;

[0037] If F s If the value is 2, the current sensor is considered to be faulty.

[0038] If F s If the value is 3, the current sensor is considered to be malfunctioning.

[0039] Preferably, the switching function δ of the k-phase bridge arm k Determined in the following manner:

[0040] The specified current flows from the rectifier to the three-phase grid side through the equivalent inductance L. k If positive, the current flows from the equivalent inductance L of the three-phase grid side. k The flow to the rectifier is negative, so the logic variable σ is defined. k , σ k =1 indicates that the k-phase current is positive, σ k =0 indicates that the k-phase current is negative;

[0041] Switch V kn The switch signal S kn The symbol "-" represents logical NOT, S kn =1 indicates that the switching transistor V kn In the on state, S kn =0 indicates that the switching transistor V kn When in the off state, the switching function δ of the k-phase bridge arm is... k The expression is:

[0042]

[0043] Preferably, the current sensor's minor fault state threshold T th1 Current sensor fault state threshold T th2 and current sensor failure threshold Tth3 The given process is as follows:

[0044] First, define the residual r4 as follows:

[0045]

[0046] Where e is the base of the natural logarithm function, r₄(0) represents the initial value of the residual r₄ at t = 0, τ is the time constant, d represents the differential, ∫ represents the first integral sign; || is the norm symbol. This is the critical value of the minute fault excitation signal ζ corresponding to the current sensor malfunction;

[0047] Current sensor minor fault state threshold T th1 Current sensor fault state threshold T th2 and current sensor failure threshold T th3 They are defined as follows:

[0048]

[0049]

[0050]

[0051] in, This represents the critical value of the minor fault excitation signal ζ corresponding to a minor fault occurring in the current sensor. This is the critical value of the minute fault excitation signal ζ corresponding to the current sensor malfunction.

[0052] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0053] 1. By using a non-singular terminal sliding mode observer and sliding mode control, the shortcomings of existing model-based methods in failing to establish accurate system models can be overcome simply by designing the sliding surface reasonably;

[0054] 2. By designing an adaptive control approach rate, the jitter characteristics in the control process are reduced and the approach rate is accelerated through adaptive switching of the approach rate, thereby further improving the accuracy of system fault diagnosis.

[0055] 3. By designing adaptive diagnostic thresholds to suppress inherent or external uncertainties in the system, diagnostic accuracy is improved;

[0056] 4. It enabled the diagnosis of minor faults in current sensors and verified the accuracy of the fault diagnosis results. Attached Figure Description

[0057] Figure 1 This is a topology diagram of the circuit involved in the embodiment of the present invention;

[0058] Figure 2 This is the control diagram of the minor fault diagnosis method of the present invention;

[0059] Figure 3 This is a flowchart of the minor fault diagnosis method of the present invention;

[0060] Figure 4 This is a simulation waveform diagram of the state variable x4 before and after the occurrence of a minor fault in an embodiment of the present invention;

[0061] Figure 5 This is a simulation waveform diagram of the diagnostic residual r4 and the three state thresholds in an embodiment of the present invention;

[0062] Figure 6 F is the fault detection feature quantity in this embodiment of the invention. s The simulation waveform diagram. Detailed Implementation

[0063] The technical solution of the present invention will now be described in further detail with reference to the accompanying drawings.

[0064] Figure 1 This is a topology diagram of the circuit involved in this embodiment of the invention. As can be seen from the diagram, the circuit topology involved in this method includes a three-phase grid-side voltage source, a three-phase grid-side equivalent inductance and a three-phase grid-side equivalent resistance, a three-phase three-level rectifier, two identical supporting capacitors, a DC-side load, a current sensor, and a control module. The two supporting capacitors are connected in series and then in parallel between the DC positive bus and the DC negative bus of the DC-side load.

[0065] The three-phase three-level rectifier is divided into three phase arms, each connected in parallel with the DC-side load. The three phase arms are denoted as k-phase arms, where k represents the phase sequence, k = a, b, c. Each phase arm consists of four switching transistors connected in series with anti-parallel diodes; therefore, the three-phase three-level rectifier contains a total of 12 switching transistors with anti-parallel diodes. These 12 switching transistors are denoted as Vswitching-transistor. kn , where n represents the serial number of the switching transistor, n = 1, 2, 3, 4. In each phase arm of the three-phase bridge, the switching transistor V k1 Switching transistor V k2 Switching transistor V k3 Switching transistor V k4 In series, the switching transistor V k2 and switching transistor V k3 The connection point is denoted as the rectifier bridge input point τ. k Let the equivalent inductance of the three-phase grid be denoted as L. k The equivalent resistance of the three-phase grid side is denoted as R. k The three-phase grid-side voltage source is denoted as E. kThe equivalent inductance L on the three-phase grid side k One end is connected to the rectifier bridge input point τ k The other end is connected to the equivalent resistance R of the three-phase grid. k The equivalent resistance R of the three-phase grid side k The other end is connected to the three-phase grid voltage source E k The positive terminal, the three-phase grid-side voltage source E k The negative terminal is grounded.

[0066] The current sensor's detection terminals are divided into three phases, denoted as detection terminals Γ. k Detection end Γ k Connected at the rectifier bridge input point τ k Equivalent inductance L of the three-phase grid side k Between them, the output terminal of the current sensor is connected to the input terminal of the control module, and the output terminal of the control module is respectively connected to 12 switching transistors V. kn .

[0067] from Figure 1 As can be seen above, each arm of the three-phase three-level rectifier also includes two clamping diodes, meaning the three-phase three-level rectifier bridge includes a total of six clamping diodes. These six clamping diodes are denoted as clamping diode D. kρ ρ is the serial number of the clamping diode, ρ = 1, 2. In each phase arm of the three-phase bridge, the clamping diode D... k1 The cathode is connected to the switching transistor V. k1 and switching transistor V k2 Between, anode clamping diode DC k2 The cathode, clamping diode D k2 The anode is connected to the switching transistor V. k3 and switching transistor V k4 Between, and clamping diode D k1 and clamping diode D k2 The connection point is connected to the midpoint O of the DC bus. Additionally... Figure 1 C1 and C2 are both supporting capacitors, R L It is a DC-side load.

[0068] In this embodiment, the equivalent inductance L on the three-phase grid side s The equivalent leakage inductance of the traction winding, the equivalent resistance R on the three-phase grid side s This is the equivalent leakage resistance of the traction winding. In this example, U a =380sin(314t), U b =380sin(314t-2π), U c =380sin(314t+2π).

[0069] Figure 2 This is a control chart for the minor fault diagnosis method of the present invention. Figure 3 This is a flowchart of the fault diagnosis method for a three-phase three-level rectifier according to the present invention. Figures 2-3 As can be seen, the steps of the diagnostic method of the present invention are as follows:

[0070] Step 1: Denote the three-phase three-level rectifier as a rectifier, and establish the hybrid logic dynamic model of the rectifier, the expression of which is: in, The phase voltage U of the k-phase bridge arm ko The estimated value, δ k For the switching function of the k-phase bridge arm.

[0071] In this embodiment, the switching function δ of the k-phase bridge arm k Determined in the following manner:

[0072] The specified current flows from the rectifier to the three-phase grid side through the equivalent inductance L. k If positive, the current flows from the equivalent inductance L of the three-phase grid side. k The flow to the rectifier is negative, so the logic variable σ is defined. k , σ k =1 indicates that the k-phase current is positive, σ k =0 indicates that the k-phase current is negative;

[0073] Switch V kn The switch signal S kn The symbol "-" represents logical NOT, S kn =1 indicates that the switching transistor V kn In the on state, S kn =0 indicates that the switching transistor V kn When in the off state, the switching function δ of the k-phase bridge arm is... k The expression is:

[0074]

[0075] Step 2, sample the equivalent resistance R flowing through the three-phase grid side. k The three-phase currents are denoted as the network-measured equivalent current i. a i b i c E at the sampling three-phase grid side voltage source k The three-phase voltage, denoted as grid-side voltage U. a U b U c Sample the voltage at the DC-side load and record it as the DC-side voltage U. dc ;

[0076] The state-space equations of the rectifier are established, and their expression is as follows:

[0077]

[0078] in, For the equivalent current i of the network measurement a i b i c The derivative of R, where R is the equivalent resistance of the three-phase grid. k The resistance value, L is the equivalent inductance of the three-phase grid. k The inductance value, D is the coefficient matrix 1, F represents the unknown disturbance signal of the rectifier.

[0079] Step 3, define the minor fault of the current sensor as minor fault f, and establish the minor fault equation, the expression of which is: in, A is the derivative of the minor fault f. f Let ξ be the Hurwitz matrix, and ξ be the excitation signal for a minor fault. In this embodiment, the value of ξ is as follows:

[0080]

[0081] Step 4: Use the state augmentation method to establish an augmented system for the small fault equations and the rectifier's state-space equations. The expression of the augmented system is as follows:

[0082]

[0083] Step 5, given the state variable x i i = 1, 2, 3, 4, state variable x i The expression is:

[0084]

[0085] The state variable x i The derivative is denoted as the state variable derivative. Then the derivative of the state variable The expressions for the augmented system output y are as follows;

[0086]

[0087] Step 6, set the state variable x i The estimated value is denoted as the state variable estimate. State variable estimates The derivative is denoted as the derivative of the state variable estimate. The estimated value of the augmented system output y is denoted as the output estimate. Construct a novel nonsingular terminal sliding mode observer, whose expression is:

[0088]

[0089]

[0090] Where g is an adjustable parameter of 1, and g > 0; k(t) is the adaptive law, and the expression for the derivative of k(t) is: t is a time variable, representing the running time of the driving system; a is the adaptive law gain coefficient. λ is an adjustable parameter 2, σ is an adjustable parameter 3, both λ and σ are positive odd numbers and satisfy 1 < λ / σ < 2, β is an adjustable parameter 4, β ∈ (0, 1); s is a non-singular terminal sliding surface. r i For residuals, Let be the derivative of the residual, and tanh() be the hyperbolic tangent function.

[0091] Step 7, define the residual r i , Then the residual r i derivative The expression is:

[0092]

[0093] Step 8, given the current sensor minor fault state threshold T th1 Current sensor fault state threshold T th2 and the current sensor failure threshold T th3 .

[0094] In this embodiment, the current sensor's minor fault state threshold T th1 Current sensor fault state threshold T th2 and the current sensor failure threshold T th3 The given process is as follows:

[0095] First, define the residual r4 as follows:

[0096]

[0097] Where e is the base of the natural logarithm function, r₄(0) represents the initial value of the residual r₄ at t = 0, τ is the time constant, d represents the differential, ∫ represents the first integral sign; || is the norm symbol. This is the critical value of the minute fault excitation signal ζ corresponding to the current sensor malfunction;

[0098] Current sensor minor fault state threshold T th1 Current sensor fault state threshold T th2 and the current sensor failure threshold T th3 They are defined as follows:

[0099]

[0100]

[0101]

[0102] in, This represents the critical value of the minor fault excitation signal ζ corresponding to a minor fault occurring in the current sensor. This is the critical value of the minute fault excitation signal ζ corresponding to the current sensor malfunction.

[0103] Step 9, Define the fault detection feature quantity F s Its expression is:

[0104] F s =(sign(||r4||-T) th1 )+1) / 2+(sign(||r4||-T th2 )+1) / 2+(sign(||r4||-T th3 )+1) / 2

[0105] Based on the fault detection characteristic quantity F s Diagnosing current sensor faults:

[0106] If F s =0, then the current sensor is considered to be in normal condition;

[0107] If F s =1, then the current sensor is considered to be in a minor fault state;

[0108] If F s If the value is 2, the current sensor is considered to be faulty.

[0109] If F s If the value is 3, the current sensor is considered to be malfunctioning.

[0110] The diagnosis is complete.

[0111] To demonstrate the beneficial effects of the present invention, a simulation was performed.

[0112] Figure 4 This is a simulation waveform diagram of the state variable x4 before and after the occurrence of a minor fault in an embodiment of the present invention. As can be seen from the figure, the state variable x4 changes when the minor fault occurs.

[0113] Figure 5 This is a simulation waveform diagram of the diagnostic residual r4 and three state thresholds in an embodiment of the present invention. As can be seen from the figure, 0.202 seconds after the minor fault occurs, r4 suddenly increases and exceeds the minor fault state threshold T of the current sensor. th1At 0.236 seconds, r4 suddenly increased and exceeded the current sensor fault state threshold T. th2 At 0.306 seconds, r4 suddenly increased and exceeded the current sensor malfunction threshold T. th3 .

[0114] Figure 6 F is the fault detection feature quantity in this embodiment of the invention. s The simulation waveform diagram is shown. As can be seen from the diagram, the fault detection characteristic quantity F... s The change from 0 to 1 at 0.202 seconds indicates a minor fault in the current sensor; the change from 1 to 2 at 0.236 seconds indicates a malfunction in the current sensor; and the change from 2 to 3 at 0.306 seconds indicates a complete failure of the current sensor.

Claims

1. A method for diagnosing minor faults in the current sensor of a three-phase three-level rectifier, wherein the circuit topology involved in the method includes a three-phase grid-side voltage source, a three-phase grid-side equivalent inductance and a three-phase grid-side equivalent resistance, a three-phase three-level rectifier, two identical supporting capacitors, a DC-side load, a current sensor, and a control module, wherein the two supporting capacitors are connected in series and then in parallel between the DC positive bus and the DC negative bus of the DC-side load; The three-phase three-level rectifier is divided into three phase arms, each connected in parallel with the DC-side load. The three phase arms are denoted as k-phase arms, where k represents the phase sequence (k = a, b, c). Each phase arm consists of four switching transistors connected in series with anti-parallel diodes. Therefore, the three-phase three-level rectifier contains a total of 12 switching transistors with anti-parallel diodes. These 12 switching transistors are denoted as Vswitching-transistor. km , n represents the serial number of the switching transistor, n = 1, 2, 3, 4; in each phase arm of the three-phase bridge, the switching transistor V k1 Switching transistor V k2 Switching transistor V k3 Switching transistor V k4 In series, among which, Switching transistor V k2 and switching transistor V k3 The connection point is denoted as the rectifier bridge input point τ. k ; Let the equivalent inductance of the three-phase grid side be denoted as L. k The equivalent resistance of the three-phase grid side is denoted as R. k The three-phase grid-side voltage source is denoted as E. k The equivalent inductance L on the three-phase grid side k One end is connected to the rectifier bridge input point τ k The other end is connected to the equivalent resistance R of the three-phase grid. k The equivalent resistance R of the three-phase grid side k The other end is connected to the three-phase grid voltage source E k The positive terminal, the three-phase grid-side voltage source E k The negative terminal is grounded; The current sensor's detection terminals are divided into three phases, denoted as detection terminals Γ. k Detection end Γ k Connected to the rectifier bridge input point τ k Equivalent inductance L of the three-phase grid side k Between them, the output terminal of the current sensor is connected to the input terminal of the control module, and the output terminal of the control module is respectively connected to 12 switching transistors V. kn ; The diagnostic method is characterized by comprising the following steps: Step 1: Denote the three-phase three-level rectifier as a rectifier, and establish the hybrid logic dynamic model of the rectifier, the expression of which is: in, The phase voltage U of the k-phase bridge arm ko The estimated value, δ k Let be the switching function of the k-phase bridge arm; Step 2, sample the equivalent resistance R flowing through the three-phase grid side. k The three-phase currents are denoted as the network-measured equivalent current i. a i b i c E at the sampling three-phase grid side voltage source k The three-phase voltage, denoted as grid-side voltage U. a U b U c Sample the voltage at the DC-side load and record it as the DC-side voltage U. dc ; The state-space equations of the rectifier are established, and their expression is as follows: in, For the equivalent current i of the network measurement a i b i c The derivative of R, where R is the equivalent resistance of the three-phase grid. k The resistance value, L is the equivalent inductance of the three-phase grid. k The inductance value, D is the coefficient matrix 1, F represents the unknown disturbance signal of the rectifier; Step 3, define the minor fault of the current sensor as minor fault f, and establish the minor fault equation, the expression of which is: in, A is the derivative of the minor fault f. f Let ξ be the Hurwitz matrix, and ξ be the excitation signal for a minor fault. Step 4: Use the state augmentation method to establish an augmented system for the small fault equations and the rectifier's state-space equations. The expression of the augmented system is as follows: Step 5, given the state variable x i i = 1, 2, 3, 4, state variable x i The expression is: The state variable x i The derivative is denoted as the state variable derivative. Then the derivative of the state variable The expressions for the augmented system output y are as follows; Step 6, set the state variable x i The estimated value is denoted as the state variable estimate. State variable estimates The derivative is denoted as the derivative of the state variable estimate. The estimated value of the augmented system output y is denoted as the output estimate. Construct a novel nonsingular terminal sliding mode observer, whose expression is: Where g is an adjustable parameter of 1, and g > 0; k(t) is the adaptive law, and the expression for the derivative of k(t) is: t is a time variable, representing the running time of the driving system; a is the adaptive law gain coefficient. λ is an adjustable parameter 2, σ is an adjustable parameter 3, both λ and σ are positive odd numbers and satisfy 1 < λ / σ < 2, β is an adjustable parameter 4, β ∈ (0, 1); s is a non-singular terminal sliding surface. r i For residuals, Let be the derivative of the residual, and tanh() be the hyperbolic tangent function. Step 7, define the residual r i , Then the residual r i derivative The expression is: Step 8, given the current sensor minor fault state threshold T th1 Current sensor fault state threshold T th2 and current sensor failure threshold T th3 ; Step 9, Define the fault detection feature quantity F s Its expression is: F s =(sign(||r4||-T th1 )+1) / 2+(sign(||r4||-T th2 )+1) / 2+(sign(||r4||-T th3 )+1) / 2 Based on the fault detection characteristic quantity F s Diagnosing current sensor faults: If F s =0, then the current sensor is considered to be in normal condition; If F s =1, then the current sensor is considered to be in a minor fault state; If F s If the value is 2, the current sensor is considered to be faulty. If F s If the value is 3, the current sensor is considered to be malfunctioning.

2. The method for diagnosing minor faults in the current sensor of a three-phase three-level rectifier according to claim 1, characterized in that, The switching function δ of the k-phase bridge arm k Determined in the following manner: The specified current flows from the rectifier to the three-phase grid side through the equivalent inductance L. k If positive, the current flows from the equivalent inductance L of the three-phase grid side. k The flow to the rectifier is negative, so the logic variable σ is defined. k , σ k =1 indicates that the k-phase current is positive, σ k =0 indicates that the k-phase current is negative; Switch V kn The switch signal S kn The symbol "-" represents logical NOT, S kn =1 indicates that the switching transistor V kn In the on state, S kn =0 indicates that the switching transistor V kn When in the off state, the switching function δ of the k-phase bridge arm is... k The expression is:

3. The method for diagnosing minor faults in the current sensor of a three-phase three-level rectifier according to claim 1, characterized in that, The current sensor's minor fault state threshold T th1 Current sensor fault state threshold T th2 and the current sensor failure threshold T th3 The given process is as follows: First, define the residual r4 as follows: Where e is the base of the natural logarithm function, r₄(0) represents the initial value of the residual r₄ at t = 0, τ is the time constant, d represents the differential, ∫ represents the first integral sign; || is the norm symbol. This is the critical value of the minute fault excitation signal ζ corresponding to the current sensor malfunction; Current sensor minor fault state threshold T th1 Current sensor fault state threshold T th2 and the current sensor failure threshold T th3 They are defined as follows: in, This represents the critical value of the minor fault excitation signal ζ corresponding to a minor fault occurring in the current sensor. This is the critical value of the minute fault excitation signal ζ corresponding to the current sensor malfunction.

Images