Multiple fault diagnosis method for permanent magnet motor drive system

Abstract

The present invention discloses a multiple fault diagnosis method for a permanent magnet motor drive system, based on d-q-x1-y1-...-x μ ‑y μ The ‑o1‑o2 current and coordinate transformation matrix is ​​used to independently decompose the current of each phase into direct-axis and non-direct-axis components. A single fault diagnosis index is designed for detecting and locating multiple single-phase open-circuit fault types, including upper tube open circuit, lower tube open circuit, simultaneous open circuit of upper and lower tubes of the same phase, or winding open circuit. The fault characteristics and fault indicators of different phases do not affect each other and can be directly linearly combined to obtain the diagnosis result of the entire system, realizing multiple open-circuit fault diagnosis of permanent magnet motor drive systems and improving the robustness to non-ideal conditions such as current imbalance and current harmonics.

Description

Technical Field

[0001] The present invention relates to a motor fault diagnosis technology, and in particular to a multiple fault diagnosis method for a permanent magnet motor drive system. Background Art

[0002] Electric drive systems are widely used in critical applications such as aerospace, ship propulsion, and high-speed train traction. Given the high reliability and safety requirements of these applications and the inherent vulnerability of various components in electric drive systems, research on fault diagnosis algorithms has significant engineering application value and theoretical research implications for the safety, reliability, and related fault-tolerant control methods of the entire electric drive system. Because most faults are often converted to open-circuit faults through hardware isolation, research on fault diagnosis algorithms has largely focused on open-circuit faults.

[0003] Currently, open-circuit fault diagnosis is mostly based on current signal analysis methods, such as the Park current vector method and harmonic current analysis. These methods directly analyze and process the current signals available in the motor drive system, eliminating the need for additional sensors and making them easy to implement. However, permanent magnet motor drive systems often suffer from non-ideal conditions such as current harmonics and current imbalance, posing challenges to the robustness and accuracy of existing open-circuit fault diagnosis methods. Summary of the Invention

[0004] Purpose of the invention: In response to the above problems, the purpose of the present invention is to provide a multiple fault diagnosis method for a permanent magnet motor drive system, to realize multiple open circuit fault diagnosis of the permanent magnet motor drive system, and to improve the robustness to non-ideal conditions such as current imbalance and current harmonics.

[0005] Technical solution: A multiple fault diagnosis method for a permanent magnet motor drive system of the present invention comprises the following steps:

[0006] Step 1, based on the coordinate transformation matrix and dq-x1-y1-...-x μ -y μ -o1-o2 current, decompose each phase current of the motor into direct axis component and non-direct axis component independently;

[0007] Step 2: pre-process the direct-axis component and non-direct-axis component of each phase current and extract fault features;

[0008] Step 3: Determine the upper and lower pipe faults based on the fault characteristics;

[0009] Step 4: Calculate the fault diagnosis index based on the upper and lower tube fault flags, and use the fault diagnosis index to diagnose multiple faults of the permanent magnet motor drive system.

[0010] Furthermore, step 1 specifically includes:

[0011] The current of each phase of the motor is converted from dq-x1-y1-...-x n -y n -o1-o2 current is expressed as follows:

[0012] [i a i b … i nm ] T =T -1 [i d i q i x1 i y1 … i xμ i yμ i o1 i o2 ] T

[0013] Where n is the number of symmetrical windings, m is the number of stator windings in each group, i a ,i b ,...,i nm is the motor phase current, i d 、i q are d-axis and q-axis currents respectively, i x1 ,i y1 ,...,i xμ ,i yμ is the harmonic current, i o1 and i o2 is the zero-sequence current; T is the coordinate transformation matrix, which is expressed as follows:

[0014]

[0015] Where P and S(μ,γ) are matrices, expressed as follows:

[0016]

[0017]

[0018] Where δ = 2π / m, γ is the phase shift between two adjacent symmetrical windings, and θ is the rotor position.

[0019] Each phase current is independently decomposed into direct axis components and non-direct axis components, which are expressed as follows:

[0020]

[0021] Furthermore, the step 2 specifically includes:

[0022] For the direct axis component i kd The sliding average calculation is expressed as follows:

[0023]

[0024] In the formula, k is any phase, t represents the current moment, T σ =σΤ f is the sliding average period, T f is the fundamental period of the phase current, σ is the ratio of the moving average period to the fundamental period of the phase current;

[0025] For non-direct axis component i knd Taking the absolute value and calculating the sliding average is expressed as follows:

[0026]

[0027] The fault features are extracted using the direct axis components and non-direct axis components after preprocessing, which are expressed as follows:

[0028] f k = kd > / <|i knd |>

[0029] When there is no fault in phase k, the fault characteristic f k =0; when an upper tube open circuit fault occurs, the fault characteristic f k =-1; when the lower tube open circuit fault occurs, the fault characteristic f k =1.

[0030] Furthermore, the step 3 specifically includes:

[0031] Determine whether the fault characteristics are near the characteristic values ​​corresponding to the upper and lower tube faults based on the following rules:

[0032] When f k ∈(-1-∈,-1+∈), let the upper tube fault characteristic f ku =1, otherwise f ku =0; where ∈ represents the hysteresis bandwidth;

[0033] When f k ∈(1-∈,1+∈), let the lower tube fault characteristic f kl =1, otherwise f kl =0;

[0034] Fault characteristics of upper and lower pipes f ku and f kl Perform moving average processing and compare it with the judgment threshold ε to judge the upper and lower tube faults respectively. The rules are as follows:

[0035] when <f ku >>ε, set the upper tube fault flag F ku =1, otherwise F ku =0;​

[0036] when <f kl >>ε, set the lower tube fault flag F kl =1, otherwise F kl =0;

[0037] Since the fault characteristics of the upper and lower tubes are filtered by the sliding average and used to determine the fault, the determination threshold is not affected by the speed change.

[0038] Furthermore, the step 4 specifically includes:

[0039] According to the upper and lower tube fault flags, the fault diagnosis indicators are as follows:

[0040] F k =F ku +2F kl

[0041] When there is no fault in phase k, F k =0; when the upper tube is open, F k =1; when the lower tube is open, F k =2; When the upper and lower tubes are open at the same time or the winding is open, F k =3, then according to F k The value of is used to obtain the k-phase fault diagnosis result.

[0042] Furthermore, the multiple faults in step 4 include upper tube open circuit, lower tube open circuit, upper and lower tubes of the same phase open circuit at the same time, winding open circuit, and upper and lower tubes of different phases open circuit at the same time.

[0043] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages:

[0044] 1. The fault diagnosis index in the present invention is always approximately zero in a healthy state, and the judgment threshold is not affected by speed changes. Therefore, it is highly robust to non-ideal factors such as sudden load changes, sudden speed changes, current harmonics, and current imbalance.

[0045] 2. The fault diagnosis index of the present invention corresponds to the number of inverter / motor phases. A single index is used to detect and locate multiple fault types, such as open circuits of the upper and lower tubes within a phase, either separately or simultaneously. The indexes of each phase do not affect each other. The fault combination between different phases can be derived from the linear combination of the indexes of each phase, which is very simple.

[0046] 3. The strong robustness of the fault diagnosis indicator in the present invention indicates that it does not make special requirements on the winding neutral point connection method, the number of motor phases, the winding structure, the harmonic plane control target, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1is a flow chart of a method for diagnosing multiple faults of a permanent magnet motor drive system in an embodiment;

[0048] Figure 2 It is the control block diagram and open circuit fault type applicable in the embodiment;

[0049] Figure 3 is a flow chart of a multiple open-circuit fault diagnosis algorithm for a permanent magnet motor drive system in an embodiment;

[0050] Figure 4 The waveforms in different states of the embodiment are as follows: (a) waveform when the upper tube is open, (b) waveform when the lower tube is open, (c) the motor phase current i before and after the upper and lower tubes are open at the same time or the winding is open k , direct axis and non-direct axis components, i kd and i knd Waveform diagram. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solutions and advantages of this application more clear, this application is further described in detail below with reference to the accompanying drawings and embodiments.

[0052] like Figure 1 The flowchart of the multiple fault diagnosis method of the permanent magnet motor drive system according to this embodiment is shown. The multiple fault diagnosis method is applicable to three-phase motors, symmetrical or asymmetrical multi-phase motors, and the winding structures correspond to the following respectively: Figure 2 Figures (a)-(c) in the middle are applicable to the following: Figure 2 There are three basic fault types, including the upper tube open circuit shown in (d), the lower tube open circuit shown in (e), and the upper and lower tubes of the same phase open circuit (or winding open circuit) shown in (f), and no special requirements are made on the harmonic plane control target, such as Figure 3 As shown, the multiple fault diagnosis method specifically includes the following steps:

[0053] Step 1, based on the coordinate transformation matrix and dq-x1-y1-...-x μ -y μ -o1-o2 current, which decomposes each phase current of the motor into direct axis component and non-direct axis component independently.

[0054] In one embodiment, the above step 1 specifically includes:

[0055] The current of each phase of the motor is converted from dq-x1-y1-...-x n -y n -o1-o2 current is expressed as follows:

[0056] [i a i b … i nm ] T=T -1 [i d i q i x1 i y1 … i xμ i yμ i o1 i o2 ] T

[0057] Where n is the number of symmetrical windings, m is the number of stator windings in each group, i a ,i b ,...,i nm is the motor phase current, i d 、i q are d-axis and q-axis currents respectively, i x1 ,i y1 ,...,i xμ ,i yμ is the harmonic current, i o1 and i o2 is the zero-sequence current; T is the coordinate transformation matrix, which is expressed as follows:

[0058]

[0059] Where P and S(μ,γ) are matrices, expressed as follows:

[0060]

[0061]

[0062] Where δ = 2π / m, γ is the phase shift between two adjacent sets of symmetrical windings, and θ is the rotor position.

[0063] Each phase current is independently decomposed into direct axis components and non-direct axis components, which are expressed as follows:

[0064]

[0065] Step 2: Preprocess the direct-axis component and non-direct-axis component of each phase current and extract fault features.

[0066] In one example, step 2 specifically includes:

[0067] In order to realize the classification of upper and lower tube open circuit faults, the direct axis component i kd The sliding average calculation is expressed as follows:

[0068]

[0069] In the formula, k is any phase, t represents the current moment, Tσ =σΤ f is the sliding average period, T f is the fundamental period of the phase current, and σ is the ratio of the moving average period to the fundamental period of the phase current. It is not affected by the speed change and can be adjusted according to the actual situation.

[0070] For non-direct axis component i knd Taking the absolute value and calculating the sliding average is expressed as follows:

[0071]

[0072] The fault features are extracted using the direct axis components and non-direct axis components after preprocessing, which are expressed as follows:

[0073] f k = kd > / <|i knd |>

[0074] When there is no fault in phase k, the fault characteristic f k =0; when an upper tube open circuit fault occurs, the fault characteristic f k =-1; when the lower tube open circuit fault occurs, the fault characteristic f k =1, as shown in Table 1, the possible open circuit fault types of phase k and their corresponding fault characteristic values. After preprocessing, f k It is not affected by speed / load changes, current harmonics, current imbalance and other factors, so it has strong robustness.

[0075] Table 1 Possible open circuit fault types of phase k and their corresponding fault characteristic values

[0076] Fault characteristic value 0 -1 1 Fault type No trouble Upper tube open Open down tube

[0077] Step 3: Determine the upper and lower pipe faults based on the fault characteristics.

[0078] In one example, step 3 specifically includes:

[0079] Determine whether the fault characteristics are near the characteristic values ​​corresponding to the upper and lower tube faults based on the following rules:

[0080] When f k ∈(-1-∈,-1+∈), let the upper tube fault characteristic f ku =1, otherwise f ku =0; where ∈ represents the hysteresis bandwidth;

[0081] When f k ∈(1-∈,1+∈), let the lower tube fault characteristic f kl =1, otherwise f kl =0; ​

[0082] Fault characteristics of upper and lower pipes f ku and f kl Perform moving average processing and compare it with the judgment threshold ε to judge the upper and lower tube faults respectively. The rules are as follows:

[0083] when <f ku >>ε, set the upper tube fault flag F ku =1, otherwise F ku =0;

[0084] when <f kl >>ε, set the lower tube fault flag F kl =1, otherwise F kl =0;

[0085] Since the fault characteristics of the upper and lower tubes are filtered by the sliding average and used to determine the fault, the determination threshold is not affected by the speed change.

[0086] Step 4: Calculate the fault diagnosis index based on the upper and lower tube fault flags, and use the fault diagnosis index to diagnose multiple faults of the permanent magnet motor drive system.

[0087] In one example, step 4 specifically includes:

[0088] According to the upper and lower tube fault flags, the fault diagnosis indicators are as follows:

[0089] F k =F ku +2F kl

[0090] When there is no fault in phase k, F k =0; when the upper tube is open, F k =1; when the lower tube is open, F k =2; When the upper and lower tubes are open at the same time or the winding is open, F k =3, then according to F k The value of is used to obtain the k-phase fault diagnosis result.

[0091] The multiple faults in step 4 above include upper tube open circuit, lower tube open circuit, upper and lower tubes of the same phase open circuit at the same time, winding open circuit, and upper and lower tubes of different phases open circuit at the same time.

[0092] Table 2 shows the possible open circuit fault types of phase k and their corresponding fault characteristic values. k The k-phase fault diagnosis result can be obtained by looking up Table 2. The fault diagnosis indicators of each phase do not affect each other, so the entire diagnosis result can be obtained by combining the fault indicators of each phase.

[0093] Table 2 Possible open circuit fault types of phase k and their fault diagnosis index values

[0094]

[0095] The program uses i in a healthy state d =0 vector control strategy.

[0096] Figure 4 The waveforms of different faults are shown: (a) upper tube open circuit, (b) lower tube open circuit, (c) upper and lower tube open circuit at the same time or the motor phase current i before and after the winding open circuit. k , direct axis and non-direct axis components (i kd and i knd ) waveform. Figure 4 It is known that after an open circuit fault occurs, the direct axis component i kd and the non-direct axis component i knd are opposite to each other. Moreover, when the upper tube is open, the direct axis component i kd The polarity is negative, such as Figure 4 (a) When the lower tube is open, the direct axis component i kd The polarity is positive, such as Figure 4 (b) When the upper and lower tubes are open at the same time or the winding is open, the direct axis component i kd When the polarity is the upper and lower tube open circuit fault, i nd Linear superposition of polarities, such as Figure 4 The above analysis shows that the fault diagnosis method can realize multiple open circuit fault diagnosis, including upper tube open circuit, lower tube open circuit, upper and lower tubes of the same phase open circuit (or winding open circuit), and upper and lower tubes of different phases open circuit.

[0097] The above merely describes a preferred embodiment of the present invention. A person skilled in the art will readily appreciate other advantages and variations based on the above embodiment. Therefore, the present invention is not limited to the above embodiment, which serves only as an example to provide a detailed, illustrative description of one form of the present invention. Any common changes and substitutions made by a person skilled in the art within the scope of the present invention's technical solution, without departing from the spirit of the present invention, should be included within the scope of protection of the present invention.

Claims

1. A multiple fault diagnosis method for a permanent magnet motor drive system, characterized in that: The steps include: Step 1, based on the coordinate transformation matrix and dq-x1-y1-...-x μ -y μ -o1-o2 current, the motor current of each phase is independently decomposed into direct axis component and non-direct axis component; specifically: The current of each phase of the motor is converted from dq-x1-y1-...-x n -y n -o1-o2 current is expressed as follows: [i a i b …i nm ] T =T -1 [i d i q i x1 i y1 …i xμ i yμ i o1 i o2 ] T Where n is the number of symmetrical windings, m is the number of stator windings in each group, i a ,i b ,...,i nm is the motor phase current, i d 、i q are d-axis and q-axis currents respectively, i x1 ,i y1 ,...,i xμ ,i yμ is the harmonic current, i o1 and i o2 is the zero-sequence current; T is the coordinate transformation matrix, which is expressed as follows: Where P and S(μ,γ) are matrices, expressed as follows: Where δ = 2π / m, γ is the phase shift between two adjacent symmetrical windings, and θ is the rotor position. Each phase current is independently decomposed into direct axis components and non-direct axis components, which are expressed as follows: Step 2: pre-process the direct-axis component and non-direct-axis component of each phase current and extract fault features; Specifically: For the direct axis component i kd The sliding average calculation is expressed as follows: In the formula, k is any phase, t represents the current moment, T σ =σΤ f is the sliding average period, T f is the fundamental period of the phase current, σ is the ratio of the moving average period to the fundamental period of the phase current; For non-direct axis component i knd Taking the absolute value and calculating the sliding average is expressed as follows: The fault features are extracted using the direct axis components and non-direct axis components after preprocessing, which are expressed as follows: f k =<i kd > / <|i knd |> When there is no fault in phase k, the fault characteristic f k =0; when an upper tube open circuit fault occurs, the fault characteristic f k =-1; when the lower tube open circuit fault occurs, the fault characteristic f k =1; Step 3: Determine the faults of the upper and lower pipes based on the fault characteristics; specifically: Determine whether the fault characteristics are near the characteristic values ​​corresponding to the upper and lower tube faults based on the following rules: When f k ∈(-1-∈,-1+∈), let the upper tube fault characteristic f ku =1, otherwise f ku =0; where ∈ represents the hysteresis bandwidth; When f k ∈(1-∈,1+∈), let the lower tube fault characteristic f kl =1, otherwise f kl =0; Fault characteristics of upper and lower pipes f ku and f kl Perform moving average processing and compare it with the judgment threshold ε to judge the upper and lower tube faults respectively. The rules are as follows: when <f ku >>ε, set the upper tube fault flag F ku =1, otherwise F ku =0; when <f kl >>ε, set the lower tube fault flag F kl =1, otherwise F kl =0; Since the fault characteristics of the upper and lower tubes are filtered by the sliding average and used to determine the fault, the determination threshold is not affected by the speed change; Step 4: Calculate the fault diagnosis index based on the upper and lower tube fault flags, and use the fault diagnosis index to diagnose multiple faults of the permanent magnet motor drive system; specifically: According to the upper and lower tube fault flags, the fault diagnosis indicators are as follows: F k =F ku +2F kl When there is no fault in phase k, F k =0; when the upper tube is open, F k =1; when the lower tube is open, F k =2; When the upper and lower tubes are open at the same time or the winding is open, F k =3, then according to F k The value of is used to obtain the k-phase fault diagnosis result.

2. The multiple fault diagnosis method of the permanent magnet motor drive system according to claim 1, characterized in that: The multiple faults in step 4 include upper tube open circuit, lower tube open circuit, upper and lower tubes of the same phase open circuit at the same time, winding open circuit, and upper and lower tubes of different phases open circuit at the same time.

Images