Method for detecting large spark of shunt excitation direct current motor

Abstract

The invention provides a method for detecting large spark of a shunt excitation direct current motor, and belongs to the technical field of motor control. The method comprises the following steps: acquiring an armature current signal and an exciting current signal of the shunt excitation direct current motor, firstly performing band-pass filtering on the armature current signal and the exciting current signal; noise interference is removed, high-frequency spark pulse signals are reserved, then wavelet analysis is carried out on the signals, the energy change rate of the high-frequency part of detail signals is used as a fault characteristic quantity, a fault diagnosis threshold value is set, real-time online detection is carried out, and the requirements of online monitoring and non-contact detection are met; the sensor is installed in a non-contact mode, data acquisition and processing processes are both carried out in a motor operation state, real-time on-line monitoring is achieved, normal operation of the motor is not affected, and extra mechanical loss caused by contact type measurement is avoided. Meanwhile, the method can identify early latent abnormity with large sparks, overcomes the problem of poor universality of an existing detection method, and is beneficial to popularization and application.

Description

Technical Field

[0001] This application belongs to the field of motor control technology, specifically relating to a method for detecting excessive sparking in a shunt-wound DC motor. Background Technology

[0002] Shunt DC motors, with their excellent speed regulation performance and starting torque, are widely used in various fields such as industrial production, transportation, and power systems. During motor operation, the commutation process between the brushes and the commutator is crucial. When the current direction in the armature winding changes, the brushes and commutator need to achieve stable current transmission to ensure the motor generates continuous and stable electromagnetic torque. However, this commutation process is easily affected by various factors, such as brush material, commutator surface condition, brush pressure, and load current. Abnormalities in these factors can lead to excessive sparking between the brushes and the commutator. Excessive sparking not only reduces motor operating efficiency and accelerates wear on the brushes and commutator, shortening motor lifespan, but can also cause serious safety accidents such as fires. Therefore, accurate detection of excessive sparking in shunt DC motors is of great significance. Existing technologies commonly use methods such as electrical signals, optical principles, and mechanical condition monitoring to detect shunt DC motors, as detailed below:

[0003] (1) Detection methods based on electrical signals: These typically include current monitoring, shaft voltage and shaft current measurement. Current monitoring involves connecting a current sensor in series in the circuit to monitor the current changes during motor operation in real time. When abnormal current fluctuations occur, it can be inferred that there may be excessive sparking. However, this method can only indirectly reflect the sparking situation and cannot accurately determine the actual size and location of the spark. Furthermore, it is easily affected by factors such as power grid harmonics and normal load fluctuations, leading to misjudgments. Shaft voltage and shaft current measurement analyzes electrical sparking activity by measuring the voltage and current on the motor shaft, and can detect internal electrical sparking. However, it requires setting up dedicated measurement points on the motor shaft, making installation and operation complex, and is significantly affected by the motor structure and electromagnetic environment.

[0004] (2) Detection methods based on optical principles: These typically include high-speed photography and photoelectric sensor detection. High-speed photography uses a high-speed camera to photograph the contact area between the brush and the commutator, and analyzes the images to determine the spark parameters, allowing for direct observation of the spark situation. However, the equipment is expensive, requires specialized image analysis software and personnel, and its installation in industrial settings is limited, making real-time online detection difficult. On the other hand, photoelectric sensor detection uses a photoelectric sensor installed inside the motor to capture the spark light signal, offering high sensitivity and response speed. However, it is easily affected by ambient light interference, requiring shielding measures that can affect detection accuracy.

[0005] (3) Mechanical condition-based detection methods: These typically include brush wear monitoring and commutator surface inspection. Brush wear monitoring uses displacement or pressure sensors mounted on the brush holder to monitor brush wear and pressure changes, allowing for early detection of potential problems. However, it cannot directly detect excessive sparking that has already occurred. Commutator surface inspection uses equipment such as laser displacement sensors or roughness measuring instruments to periodically inspect the commutator surface condition. It can directly detect the commutator's condition, but it requires machine downtime for inspection, affecting the motor's normal operating time and making it unsuitable for continuous production scenarios.

[0006] In practical applications, the existing detection methods mentioned above all have their limitations and shortcomings. Therefore, it is necessary to propose a new method for detecting excessive spark in shunt-wound DC motors to solve the problems of existing detection methods, such as difficulty in accurately locating the source of sparks, difficulty in real-time monitoring, and poor versatility. Summary of the Invention

[0007] In view of the technical problems existing in the background art, this application provides a method for detecting excessive spark in a shunt-wound DC motor, which solves the problems of existing detection methods being unable to accurately locate the source of sparks, being difficult to monitor in real time, and having poor versatility.

[0008] In a first aspect, embodiments of this application provide a method for detecting excessive sparking in a shunt-wound DC motor, comprising the following steps:

[0009] S1 acquires the current and speed signals of the shunt-wound DC motor without stopping the machine, performs bandpass filtering on the current signal to obtain high-frequency current filtered data, and performs low-frequency filtering on the speed signal to obtain low-frequency speed filtered data; wherein, the current signal includes armature current signal and excitation current signal.

[0010] S2, perform wavelet decomposition on the current filtered data to obtain real-time current detail signal data and real-time current approximate signal data, calculate the energy change rate of the real-time current detail signal data relative to the current detail signal data under normal operation, and determine the fault characteristic quantity of excessive spark and the current fluctuation amplitude based on the energy change rate and the preset fault diagnosis threshold.

[0011] S3, the changing trend of speed filter data and the current fluctuation amplitude are used to comprehensively judge whether there is a fault of excessive sparking in the shunt-wound DC motor.

[0012] In some embodiments, step S1 specifically includes the following steps:

[0013] S11 uses a Hall current sensor to collect the armature current signal and excitation current signal of the shunt DC motor, and uses a non-contact signal acquisition system to collect the speed signal of the shunt DC motor.

[0014] S12 performs bandpass filtering on the armature current signal and the excitation current signal to obtain high-frequency armature current filtered data and excitation current filtered data, and performs low-frequency filtering on the speed signal to obtain low-frequency speed filtered data.

[0015] In some embodiments, step S2 specifically includes the following steps:

[0016] S21 uses the db8 wavelet function to perform wavelet decomposition on the armature current filtered data and the excitation current filtered data, respectively obtaining real-time armature current detail signal data, real-time armature current approximate signal data, real-time excitation current detail signal data and real-time excitation current approximate signal data composed of different frequency layers.

[0017] S22, calculate the energy corresponding to each layer of the real-time armature current detail signal data and the real-time armature current approximate signal data, and obtain its first energy change rate ΔW relative to the energy corresponding to the armature current detail signal data and the approximate signal data under normal operating conditions. j The energy corresponding to each layer of the real-time excitation current detailed signal data and the real-time excitation current approximate signal data is calculated, and the second energy change rate ΔQ is obtained relative to the energy corresponding to the excitation current detailed signal data and the approximate signal data under normal operating conditions. j .

[0018] S23, the preset fault diagnosis threshold for the armature current corresponding to the j-th layer is ΔW'. j The fault diagnosis threshold for excitation current is ΔQ' j Satisfying ΔW j >ΔW' j or ΔQ j >ΔQ' j When the armature current data or excitation current data of the j-th layer are determined as the characteristic quantity of excessive spark fault, j is a positive integer and j is less than or equal to the wavelet decomposition layer number.

[0019] It should be noted that in the actual calculation process, since the data that can reflect the difference between the two operating conditions of normal operation and excessive sparking of the shunt DC motor mainly comes from the detailed signal data, regardless of whether it is the armature current or the excitation current, after wavelet decomposition and obtaining the corresponding detailed signal data and approximate signal data components, the approximate signal data component can be ignored. Only the energy change rate of the detailed signal data component under the two operating conditions can be considered to determine the fault characteristic quantity of excessive sparking, and thus determine whether excessive sparking has occurred.

[0020] Furthermore, in some embodiments, the energy corresponding to each layer of the real-time armature current detail signal data... The calculation formula is:

[0021] (1)

[0022] Real-time excitation current detail signal data and the corresponding energy of each layer The calculation formula is:

[0023] (2)

[0024] First rate of energy change ΔW j The calculation formula is:

[0025] (3)

[0026] Second energy change rate ΔQ j The calculation formula is:

[0027] (4)

[0028] in, This represents the energy corresponding to the j-th layer of armature current detail signal data under normal operating conditions. N represents the energy corresponding to the j-th layer of the detailed signal data of the electromagnetic excitation current under normal operating conditions. j Let be the length of the detail coefficients of the j-th layer, k represent the k-th data of the wavelet decomposition detail signal of the j-th layer, and d represent the detail signal.

[0029] In some embodiments, step S23 specifically includes the following steps: the fault diagnosis threshold ΔW' of the armature current corresponding to the j-th layer. j The fault diagnosis threshold ΔQ' of the excitation current corresponding to the j-th layer is 40%. j It is 40%.

[0030] In some embodiments, step S3 specifically includes the following steps:

[0031] S31, based on the energy corresponding to each layer of real-time armature current detail signal data. Real-time excitation current detail signal data and the corresponding energy of each layer During normal operation, the energy corresponding to each layer of the armature current detail signal data is: The energy corresponding to each layer of the excitation current detailed signal data during normal operation is ,Sure and relative to and The nth layer, where the overall energy change is greatest, is the observation layer.

[0032] S32, observe the up-and-down vibration of the wavelet in the observation layer within i periods, count the actual current fluctuation ratio of the number of times the up-and-down vibration exceeds the vibration threshold within i periods, compare the actual current fluctuation ratio with the preset safe fluctuation ratio, and obtain the current fluctuation amplitude.

[0033] S33, considering the changes in the overall speed filter data and the current fluctuation amplitude, if any of the following conditions are met, then the shunt-wound DC motor is judged to have an excessive spark fault; otherwise, there is no excessive spark fault:

[0034] A1) The waveform change trend of the speed filtering data is not that it gradually decreases first and then tends to stabilize and oscillate slightly;

[0035] A2) The actual current fluctuation ratio is higher than the safe fluctuation ratio, that is, the current fluctuation amplitude is too large;

[0036] Where n and i are both positive integers.

[0037] In some embodiments, step S32 specifically includes the following steps:

[0038] S321, observe the up-and-down oscillation of the wavelet in the observation layer within i periods, count the number of times the up-and-down oscillation exceeds the oscillation threshold within i periods, and calculate the actual current fluctuation ratio Z1.

[0039] S322, preset safety fluctuation ratio Z0, compare the actual current fluctuation ratio Z1 with the preset safety fluctuation ratio Z0 to obtain the current fluctuation amplitude;

[0040] If Z1>Z0, the current fluctuation amplitude is too large; otherwise, the current fluctuation amplitude is normal. The actual current fluctuation ratio Z1=m / i, where m is a positive integer.

[0041] Furthermore, in some embodiments, Z0=20% is preferred, while in other embodiments, an appropriate safety fluctuation ratio can be preset according to the actual situation, which is not limited here.

[0042] Secondly, embodiments of this application provide a device for detecting excessive spark in a shunt-wound DC motor. This device is used to implement the method for detecting excessive spark in a shunt-wound DC motor described in the first aspect. Specifically, it includes a data acquisition box and an analysis system. The signal input terminal of the data acquisition box is connected to the shunt-wound DC motor under test and is used to acquire the current signal and speed signal of the shunt-wound DC motor without stopping the motor. The current signal and speed signal are then filtered to obtain current-filtered data and speed-filtered data, respectively. The signal input terminal of the analysis system is connected to the signal output terminal of the data acquisition box and is used to receive and analyze the current-filtered data and speed-filtered data from the data acquisition box, and to comprehensively determine whether the shunt-wound DC motor has an excessive spark fault.

[0043] Furthermore, in some embodiments, the acquisition box includes a current acquisition module and a speed acquisition module; the current acquisition module is connected to the shunt DC motor under test, and is used to acquire the current signal of the shunt DC motor and perform high-frequency filtering on the current signal to obtain high-frequency current filtered data; the speed acquisition module is connected to the shunt DC motor under test, and is used to acquire the speed signal of the shunt DC motor and perform low-frequency filtering on the speed signal to obtain low-frequency speed filtered data.

[0044] Furthermore, in some embodiments, the analysis system includes a current signal conditioning module, a current signal analysis module, a speed signal analysis module, and a comprehensive analysis module. The input of the current signal conditioning module is connected to the output of the current acquisition module, and is used to receive current filtering data, perform wavelet decomposition, and obtain real-time current detail signal data. The input of the current signal analysis module is connected to the output of the current signal conditioning module, and is used to calculate the energy change rate of the real-time current detail signal data relative to the current detail signal data under normal operating conditions. Based on the energy change rate and a preset fault diagnosis threshold, the characteristic quantity of excessive spark fault and the current fluctuation amplitude are determined. The input of the speed signal analysis module is connected to the output of the speed acquisition module, and is used to receive speed filtering data and determine the waveform change state of the speed filtering data. The input of the comprehensive analysis module is connected to the outputs of both the current signal analysis module and the speed signal analysis module, and is used to integrate the waveform change state and current fluctuation amplitude of the speed filtering data to determine whether the shunt-wound DC motor has an excessive spark fault.

[0045] Compared with the prior art, the beneficial effects of this application include:

[0046] 1. This application collects the armature current signal and excitation current signal of a shunt-wound DC motor. First, the signals are filtered to remove noise interference. Wavelet analysis is then performed on the armature current and excitation current to focus on the energy changes of the detailed signals. The energy change rate of the high-frequency part of the detailed signals is used as the fault characteristic quantity to set the fault diagnosis threshold and perform real-time online detection, meeting the requirements of online monitoring and non-contact detection. The sensor adopts non-contact installation, and the data acquisition and processing are carried out while the motor is running, realizing real-time online monitoring without affecting the normal operation of the motor and avoiding the additional mechanical losses introduced by contact measurement.

[0047] 2. This application mainly focuses on real-time online monitoring and data processing of the armature current and excitation current of a shunt-wound DC motor. Compared with mechanical signals, electrical signals are more sensitive and can identify early latent anomalies such as weak sparks, small changes in current and vibration. By reasonably setting diagnostic thresholds, early warning and intervention of faults can be achieved, extending the service life of the motor and ensuring its safe and stable operation.

[0048] 3. The detection method of this application is relatively universal and can be applied to shunt-wound DC motors with different power, speed and application scenarios. It overcomes the problem of poor universality of existing detection methods and is conducive to promotion and application. Attached Figure Description

[0049] To more clearly illustrate the technical solutions of this application, the accompanying drawings used in this application will be briefly described below. Obviously, the drawings described below are merely some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without any creative effort.

[0050] Figure 1 This is a flowchart illustrating one embodiment of the method for detecting excessive spark in a shunt-wound DC motor provided in this application.

[0051] Figure 2 A detailed flowchart illustrating one embodiment of the method for detecting excessive spark in a shunt-wound DC motor provided in this application;

[0052] Figure 3 A schematic diagram of one embodiment of the shunt-wound DC motor spark over-excess detection device provided in this application;

[0053] Figure 4 This is a flowchart illustrating the excessive spark detection device for a shunt-wound DC motor in Embodiment 1 of this application.

[0054] Figure 5 This is a schematic diagram of the shunt-wound DC motor spark detection device in Embodiment 1 of this application;

[0055] Figure 6a This is a waveform diagram of the armature current of the shunt-wound DC motor during normal operation in Embodiment 1 of this application;

[0056] Figure 6b This is a waveform diagram of the excitation current of the shunt-wound DC motor during normal operation in Embodiment 1 of this application;

[0057] Figure 6c This is a waveform diagram of the rotational speed of the shunt-wound DC motor during normal operation in Embodiment 1 of this application;

[0058] Figure 7a This is a waveform diagram of the armature current when the spark of the shunt-wound DC motor is too large in Embodiment 1 of this application;

[0059] Figure 7b This is a waveform diagram of the excitation current when the spark of the shunt-wound DC motor is too large in Embodiment 1 of this application;

[0060] Figure 7c This is a waveform diagram of the rotational speed of the shunt-wound DC motor in Embodiment 1 of this application when the spark is too large;

[0061] Figure 8a This is a waveform analysis result of the armature current during normal operation of the shunt-wound DC motor in Embodiment 1 of this application;

[0062] Figure 8b This is a waveform analysis result of the excitation current during normal operation of the shunt-wound DC motor in Embodiment 1 of this application;

[0063] Figure 8c This is a waveform analysis result of the armature current when the spark of the shunt-wound DC motor is too large in Embodiment 1 of this application;

[0064] Figure 8d This is a waveform analysis result of the excitation current when the spark of the shunt-wound DC motor is too large in Embodiment 1 of this application. Detailed Implementation

[0065] The embodiments of the technical solution of this application will now be described in detail with reference to the accompanying drawings. These embodiments are only used to more clearly illustrate the technical solution of this application and are therefore merely examples, and should not be used to limit the scope of protection of this application.

[0066] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0067] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0068] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0069] In the description of the embodiments of this application, the term "several" refers to two or more (including two), and "multiple sheets" refers to two or more (including two).

[0070] Existing technologies typically employ methods such as electrical signals, optical principles, and mechanical condition monitoring to detect shunt-wound DC motors. However, in practical applications, the commutation process of the brushes and commutator during motor operation is influenced by a combination of electromagnetic, mechanical, and thermal factors (such as sudden load changes, brush wear, and commutator surface defects). Existing detection methods often face the following main problems:

[0071] 1) Difficulty in real-time tracking of dynamic processes: The generation of commutation sparks is instantaneous (millisecond level). Traditional single-point monitoring (such as a single photoelectric sensor or current sensor) is difficult to capture the dynamic evolution process of the spark (such as spark frequency and energy distribution), resulting in delayed fault location.

[0072] 2) Insufficient detection accuracy and reliability: Traditional detection methods are easily affected by external environmental factors, leading to misjudgment and missed judgment, which cannot provide reliable protection for the safe and stable operation of motors.

[0073] 3) Difficult to meet the needs of online monitoring and non-contact testing: Most existing testing methods require shutdown testing or use contact measurement, which cannot monitor in real time during motor operation. Furthermore, contact measurement may introduce additional mechanical losses, affecting motor performance.

[0074] 4) Insufficient early warning capability: Existing technologies mostly focus on post-event detection (such as alarming only when sparks are clearly visible), making it difficult to identify early latent anomalies such as excessive sparks (such as weak sparks or abnormally high brush contact resistance). This makes it impossible to intervene in faults in advance, leading to an increased risk of motor failure. Excessive sparks are often an "intermediate stage" of motor failure. For example, it may slowly develop from slight brush wear or commutator oxidation in the early stages. If it is not detected in time, it may gradually evolve into serious faults such as commutator burning or brush breakage.

[0075] 5) Poor versatility: Existing testing methods are mostly designed for specific motor models and lack adjustable parameter adaptation mechanisms, which cannot meet the testing requirements of different models of shunt DC motors. Furthermore, due to differences in power, speed, and application scenarios (such as industrial drive motors and rail transit traction motors), the threshold and characteristics of large sparks of shunt DC motors vary significantly. This leads to a significant decrease in the accuracy of existing testing methods when applied across different models.

[0076] To address the aforementioned problems, this application provides a method for detecting excessive sparking in a shunt-wound DC motor. This method acquires the armature current and excitation current signals of the shunt-wound DC motor. First, the signals are filtered to remove noise interference. Then, wavelet analysis is performed on the armature current and excitation current to focus on the energy changes of detailed signals. The energy change rate of the high-frequency components of the detailed signals is used as a fault characteristic quantity to set a fault diagnosis threshold, enabling real-time online detection that meets the requirements of online monitoring and non-contact detection. The sensor is installed non-contactly, and data acquisition and processing are performed while the motor is running, achieving real-time online monitoring without affecting normal motor operation and avoiding the additional mechanical losses introduced by contact measurements.

[0077] like Figure 1 As shown, in a first aspect, this application provides a method for detecting excessive sparking in a shunt-wound DC motor, comprising the following steps:

[0078] S1: Acquire the current and speed signals of the shunt-wound DC motor without shutting down. Perform bandpass filtering on the current signal to obtain high-frequency current filtered data, and perform low-frequency filtering on the speed signal to obtain low-frequency speed filtered data. In this step, the current signal includes both armature current and excitation current signals. Compared to the single current signal in existing technologies, this provides richer data samples and dimensions, thereby improving the stability of subsequent detection and judgment.

[0079] In some embodiments, step S1 specifically includes the following steps:

[0080] S11, a Hall current sensor is used to acquire the armature current signal and excitation current signal of the shunt-wound DC motor, and a non-contact signal acquisition system is used to acquire the speed signal of the shunt-wound DC motor. In this step, a non-contact, anti-interference, and highly isolated measuring device is used to measure the armature current, excitation current, and speed signals of the DC motor with high precision without stopping the DC motor. In other embodiments, other signal acquisition devices with the same function can be selected according to the actual situation, and are not limited here.

[0081] S12: Bandpass filtering is performed on the armature current signal and the excitation current signal to obtain high-frequency armature current filtered data and excitation current filtered data. Low-frequency filtering is performed on the speed signal to obtain low-frequency speed filtered data. High-frequency filtering is used for the armature current signal and the excitation current signal in this step because experiments have shown that during normal operation, the armature current and excitation current of a DC motor are mainly composed of DC components, with relatively few high-frequency components. When the spark is excessive, the high-frequency components in the armature current and excitation current increase significantly. Therefore, this serves as a reference indicator for judging a DC motor with excessive spark. Thus, high-frequency filtering is performed on the armature current signal and the excitation current signal before wavelet decomposition to retain the high-frequency spark pulse signal for subsequent wavelet analysis. The speed signal mainly reflects changes in its mechanical signal, which has a slower response speed than electrical signals. Therefore, low-frequency filtering is required to retain the low-frequency speed filtered data. In other embodiments, the different components in the filtering process can be specifically set according to the actual situation, which is not limited here.

[0082] S2, the current filtered data is decomposed using wavelet decomposition to obtain real-time current detail signal data and real-time current approximate signal data. The energy change rate of the real-time current detail signal data relative to the current detail signal data under normal operating conditions is calculated. Based on the energy change rate and a preset fault diagnosis threshold, the fault characteristic quantities of excessive sparking and the current fluctuation amplitude are determined. In this step, wavelet decomposition of the current filtered data is used, which can simultaneously perform multi-scale analysis of the signal in the time and frequency domains. When the signal undergoes a sudden change, the coefficients after wavelet decomposition have a modulus maxima. Therefore, the time of fault occurrence can be determined by detecting the modulus maxima. When a fault occurs in a DC motor, the current signal in the armature winding often contains a large number of time-varying, short-term impulse, and sudden components. Traditional signal analysis methods, such as Fourier transform, cannot detect sudden signals and cannot effectively extract the fault characteristic quantities of the DC motor. However, for non-stationary signals, wavelet decomposition of the original signal has the ability to characterize the local features of the signal in both the time and frequency domains, which is significantly better than traditional methods. This enables the online fault diagnosis system for DC motors and improves the accuracy of motor parameter testing.

[0083] In some embodiments, step S2 specifically includes the following steps:

[0084] S21 uses the db8 wavelet function to perform wavelet decomposition on the armature current filtered data and the excitation current filtered data, respectively obtaining real-time armature current detail signal data, real-time armature current approximate signal data, real-time excitation current detail signal data and real-time excitation current approximate signal data composed of different frequency layers.

[0085] S22, calculate the energy corresponding to each layer of the real-time armature current detail signal data and the real-time armature current approximate signal data, and obtain its first energy change rate ΔW relative to the energy corresponding to the armature current detail signal data and the approximate signal data under normal operating conditions. j The energy corresponding to each layer of the real-time excitation current detailed signal data and the real-time excitation current approximate signal data is calculated, and the second energy change rate ΔQ is obtained relative to the energy corresponding to the excitation current detailed signal data and the approximate signal data under normal operating conditions. j .

[0086] S23, the preset fault diagnosis threshold for the armature current corresponding to the j-th layer is ΔW'. j The fault diagnosis threshold for excitation current is ΔQ' j Satisfying ΔW j >ΔW' j or ΔQ j >ΔQ' j At that time, the armature current data or excitation current data of the j-th layer are determined as the characteristic quantity of excessive spark fault; where j is a positive integer, and j is less than or equal to the wavelet decomposition layer number; the fault diagnosis threshold ΔW' of the armature current corresponding to the j-th layer. j The preferred value is 40%, and the fault diagnosis threshold for the excitation current corresponding to the j-th layer is ΔQ'. j The preferred value is 40%.

[0087] Furthermore, in some embodiments, the energy corresponding to each layer of the real-time armature current detail signal data... The calculation formula is:

[0088] (1)

[0089] Real-time excitation current detail signal data and the corresponding energy of each layer The calculation formula is:

[0090] (2)

[0091] First rate of energy change ΔW j The calculation formula is:

[0092] (3)

[0093] Second energy change rate ΔQ j The calculation formula is:

[0094] (4)

[0095] in, This represents the energy corresponding to the j-th layer of armature current detail signal data under normal operating conditions. N represents the energy corresponding to the j-th layer of the detailed signal data of the electromagnetic excitation current under normal operating conditions. j Let be the length of the detail coefficients of the j-th layer, k represent the k-th data of the wavelet decomposition detail signal of the j-th layer, and d represent the detail signal.

[0096] S3, the changing trend of the speed filter data and the current fluctuation amplitude are used to comprehensively determine whether there is an excessive spark fault in the shunt-wound DC motor. In this step, both speed and current data are combined to determine whether there is an excessive spark fault. Compared with the existing technology that only uses speed or current data for judgment, it can obtain richer data dimensions, thereby improving the stability and accuracy of the judgment.

[0097] In some embodiments, step S3 specifically includes the following steps:

[0098] S31, based on the energy corresponding to each layer of real-time armature current detail signal data. Real-time excitation current detail signal data and the corresponding energy of each layer During normal operation, the energy corresponding to each layer of the armature current detail signal data is: The energy corresponding to each layer of the excitation current detailed signal data during normal operation is and relative to and The nth layer, where the overall energy change is greatest, is the observation layer.

[0099] S32: Observe the wavelet's vertical oscillation within i periods of the observation layer, count the actual current fluctuation ratio of the number of times the vertical oscillation exceeds the oscillation threshold within i periods, and compare the actual current fluctuation ratio with the preset safe fluctuation ratio to obtain the current fluctuation amplitude. Further, step S32 specifically includes the following steps:

[0100] S321, observe the up-and-down vibration of the wavelet in the observation layer within i periods, count the number of times m the up-and-down vibration exceeds the vibration threshold within i periods, and calculate the actual current fluctuation ratio Z1.

[0101] S322, a preset safety fluctuation ratio Z0 is established. The actual current fluctuation ratio Z1 is compared with the preset safety fluctuation ratio Z0 to obtain the current fluctuation amplitude. If Z1 > Z0, the current fluctuation amplitude is too large; conversely, the current fluctuation amplitude is normal. The actual current fluctuation ratio Z1 = m / i, where m is a positive integer. Further, in some embodiments, Z0 = 20% is preferred. In other embodiments, a suitable safety fluctuation ratio can be preset according to actual conditions, and this is not limited here.

[0102] S33, considering the changes in the overall speed filter data and the current fluctuation amplitude, if any of the following conditions are met, then the shunt-wound DC motor is judged to have an excessive spark fault; otherwise, there is no excessive spark fault:

[0103] A1) The waveform change trend of the speed filtering data is not that it gradually decreases first and then tends to stabilize and oscillate slightly;

[0104] A2) The actual current fluctuation ratio is higher than the safe fluctuation ratio, that is, the current fluctuation amplitude is too large;

[0105] Where n and i are both positive integers.

[0106] like Figure 2 As shown, in a second aspect, embodiments of this application provide a device for detecting excessive spark in a shunt-wound DC motor. This device is used to implement the method for detecting excessive spark in a shunt-wound DC motor described in the first aspect. Specifically, it includes a data acquisition box and an analysis system. The signal input terminal of the data acquisition box is connected to the shunt-wound DC motor under test and is used to acquire the current signal and speed signal of the shunt-wound DC motor without stopping the motor. The current signal and speed signal are then filtered to obtain current-filtered data and speed-filtered data, respectively. The signal input terminal of the analysis system is connected to the signal output terminal of the data acquisition box and is used to receive and analyze the current-filtered data and speed-filtered data from the data acquisition box, and to comprehensively determine whether the shunt-wound DC motor has an excessive spark fault.

[0107] Furthermore, in some embodiments, the acquisition box includes a current acquisition module and a speed acquisition module; the current acquisition module is connected to the shunt DC motor under test, and is used to acquire the current signal of the shunt DC motor and perform high-frequency filtering on the current signal to obtain high-frequency current filtered data; the speed acquisition module is connected to the shunt DC motor under test, and is used to acquire the speed signal of the shunt DC motor and perform low-frequency filtering on the speed signal to obtain low-frequency speed filtered data.

[0108] Furthermore, in some embodiments, the analysis system includes a current signal conditioning module, a current signal analysis module, a speed signal analysis module, and a comprehensive analysis module. The input of the current signal conditioning module is connected to the output of the current acquisition module, and is used to receive current filtering data, perform wavelet decomposition, and obtain real-time current detail signal data. The input of the current signal analysis module is connected to the output of the current signal conditioning module, and is used to calculate the energy change rate of the real-time current detail signal data relative to the current detail signal data under normal operating conditions. Based on the energy change rate and a preset fault diagnosis threshold, the characteristic quantity of excessive spark fault and the current fluctuation amplitude are determined. The input of the speed signal analysis module is connected to the output of the speed acquisition module, and is used to receive speed filtering data and determine the waveform change state of the speed filtering data. The input of the comprehensive analysis module is connected to the outputs of both the current signal analysis module and the speed signal analysis module, and is used to integrate the waveform change state and current fluctuation amplitude of the speed filtering data to determine whether the shunt-wound DC motor has an excessive spark fault.

[0109] The following are some specific embodiments. It should be noted that the embodiments described below are exemplary and are only used to explain this application, and should not be construed as limiting this application.

[0110] Example 1

[0111] Please see Figure 4 and Figure 5 The specific steps of the method for detecting excessive sparking in a shunt-wound DC motor in this embodiment are as follows:

[0112] 1) First, install the high-precision open-type Hall current sensor on the armature bus and excitation bus of the DC motor to collect the armature current and excitation current signals of the DC motor.

[0113] 2) Connect the DC motor armature current signal and excitation current signal to the signal conditioning board. The signal conditioning board is equipped with a bandpass filter to perform bandpass filtering on the collected DC motor armature current and excitation current signals, filtering out strong low-frequency background noise (such as motor main signal, power frequency interference, etc.) and retaining or highlighting high-frequency pulse signals.

[0114] 3) Convert the filtered DC motor armature current signal and excitation current signal into digital signals, and acquire the filtered DC motor armature current signal and excitation current signal through the data acquisition system to obtain real-time current detail signal data.

[0115] 4) Connect the collected real-time current detail signal data to the host computer, and use MATLAB software to perform wavelet decomposition on the DC motor armature current signal and excitation current signal. Use the db8 wavelet to perform 6 decompositions to obtain the D1~D6 layer signals.

[0116] 5) Display the waveforms of the DC motor armature current and excitation current in real time, and display the waveforms of each signal in the wavelet decomposition layer of the DC motor armature current and excitation current in real time. Calculate the energy change before and after each detail signal layer. Specifically, calculate the energy corresponding to each layer of real-time armature current detail signal data and real-time armature current approximate signal data, and obtain the first energy change rate ΔW relative to the energy corresponding to the armature current detail signal data and approximate signal data under normal operating conditions. j The energy corresponding to each layer of the real-time excitation current detailed signal data and the real-time excitation current approximate signal data is calculated, and the second energy change rate ΔQ is obtained relative to the energy corresponding to the excitation current detailed signal data and the approximate signal data under normal operating conditions. j Set the fault diagnosis threshold ΔW' for armature current. j The fault diagnosis threshold for the excitation current corresponding to the j-th layer is ΔQ', which is 40%. j 40%, ΔW j With ΔW' j ΔQ j With ΔQ' j By comparing the values, the fault characteristic quantity of excessive spark was determined.

[0117] 6) Energy corresponding to each layer based on real-time armature current detail signal data Real-time excitation current detail signal data and the corresponding energy of each layer During normal operation, the energy corresponding to each layer of the armature current detail signal data is: The energy corresponding to each layer of the excitation current detailed signal data during normal operation is and relative to and The nth layer, where the overall energy change is greatest, is the observation layer.

[0118] 7) First, the speed pulse of the DC motor is acquired by a laser tachometer, and then the analog signal of the DC motor speed is obtained by a pulse shaping circuit. Then, the analog speed signal is converted into a digital speed signal by a timer, and the DC motor speed waveform is acquired by a data acquisition system. The acquired data is then uploaded to the host computer to display the waveform trend of the DC motor speed in real time.

[0119] 8) Based on the waveform trend chart, observe the up-and-down oscillations of the wavelet within i periods of the observation layer, count the number of times m the up-and-down oscillations exceed the oscillation threshold within i periods, and calculate the actual current fluctuation ratio Z1. A preset safe fluctuation ratio Z0 = 20% is used. Compare the actual current fluctuation ratio Z1 with the preset safe fluctuation ratio Z0 to obtain the current fluctuation amplitude. If Z1 > Z0, the current fluctuation amplitude is too large; otherwise, the current fluctuation amplitude is normal. The actual current fluctuation ratio Z1 = m / i, where m is a positive integer.

[0120] 9) Based on the changes in the overall speed filter data and the current fluctuation amplitude, if any of the following conditions are met, the shunt-wound DC motor is judged to have an excessive spark fault; otherwise, it is operating normally:

[0121] A1) The waveform change trend of the speed filtering data is not that it gradually decreases first and then tends to stabilize and oscillates slightly;

[0122] A2) The actual current fluctuation ratio is higher than the safe fluctuation ratio, that is, the current fluctuation amplitude is too large.

[0123] Test Results and Analysis

[0124] This implementation method mainly focuses on testing under two operating conditions: normal operation and excessive sparking, to verify the feasibility of the method.

[0125] First, the armature current and excitation current of the DC motor are acquired, and bandpass filtering is applied to both to remove strong low-frequency background noise (such as the motor's main signal and power frequency interference) while preserving or highlighting the high-frequency spark pulse signal. Specifically, the armature current waveform, excitation current waveform, and speed waveform of the shunt-wound DC motor during normal operation are shown below. Figures 6a-6c As shown; the armature current waveform, excitation current waveform, and speed waveform of a shunt-wound DC motor when the spark is too large are as follows: Figures 7a-7c As shown.

[0126] Then, wavelet decomposition was performed on the armature current and excitation current of the DC motor respectively. The db8 wavelet function was used, and six decompositions were performed. The sampling time was 20 seconds, the sampling frequency was 100 Hz, and a total of 2000 discrete points were collected. The energy of each level of detail components was calculated and compared with the data during normal operation of the DC motor. Specifically, the wavelet analysis results of the armature current during normal operation of the shunt-wound DC motor are as follows: Figure 8a As shown, the wavelet analysis results of the excitation current during normal operation of the shunt-wound DC motor are as follows: Figure 8b As shown, the wavelet analysis results of the armature current when the spark of the shunt-wound DC motor is too large are as follows: Figure 8c As shown, the wavelet analysis results of the excitation current when the spark of the shunt-wound DC motor is too large are as follows: Figure 8dAs shown in Table 1, the corresponding energy of the armature current and its changes under normal operation and excessive spark conditions are shown in Table 2.

[0127] Table 1. Comparison of Armature Current and Corresponding Energy and Their Variation under Different Operating Conditions

[0128] Signal layer name Armature current energy during normal operation Armature current energy when spark is too large Larger sparks relative to normal operating energy variation coefficient A6 6998000 6764200 0.966 D1 34.23 41.78 1.22 D2 308.5 392.57 1.27 D3 1634.1 2320.6 1.42 D4 23279 23302 1 D5 3804.2 3836.6 1 D6 1151.5 1180.1 1.02

[0129] Table 2 Comparison of Excitation Current and its Variation under Different Operating Conditions

[0130] Signal layer name Excitation current energy during normal operation Excitation current energy when spark is too large Larger sparks relative to normal operating energy variation coefficient A6 8395.4 8071.2 0.961 D1 0.038 0.046 1.21 D2 0.343 0.437 1.27 D3 1.80 2.567 1.43 D4 26.05 26.08 1 D5 4.284 4.321 1 D6 1.372 1.405 1.02

[0131] Combination Figures 6a to 8d Based on the data shown in Tables 1 and 2, without considering approximate components, the armature current and excitation current have the largest proportion of wavelet energy in the D4 layer among the detailed components. The energy changes of the wavelet decomposition signals of the armature current and excitation current in the D1, D2, and D3 layers are relatively large, that is, the energy of the mid-to-high frequency part increases by more than 20%, and the energy change of the wavelet decomposition signal in the D3 layer is the largest, exceeding 40%. Therefore, the proportion of wavelet energy in the D4 layer of the armature current and excitation current, and the changes in the energy of the wavelet decomposition signals in the D1, D2, and D3 layers can be used to determine whether the DC motor has a fault of excessive sparking. Moreover, the energy change coefficient has a maximum value, and the detection of the maximum value point can be used to determine the time of fault occurrence. Combined with the fact that the DC motor speed first decreases and then gradually tends to stabilize with slight fluctuations, this can be used to determine whether the DC motor has a fault of excessive sparking. That is, it verifies that the detection method described in this application can effectively detect and determine the two operating conditions of the shunt-wound DC motor under normal operation and excessive sparking without stopping the machine.

[0132] In summary, compared with existing identification technologies, the method described in this application can not only ensure the identification accuracy and stability of rice planting areas without the need for actual test samples, but also achieve efficient identification of rice planting areas under complex environmental conditions such as cloudy / partly cloudy conditions.

[0133] It should be noted that this application is not limited to the above-described embodiments. The above embodiments are merely examples, and any embodiments with the same structure and effect as the technical concept within the scope of this application are included in the technical scope of this application. Furthermore, various modifications that can be conceived by those skilled in the art to the embodiments, and other ways of constructing by combining some of the constituent elements of the embodiments, without departing from the spirit of this application, are also included in the scope of this application.

Claims

1. A method for detecting spark overreach of a shunt DC motor, characterized by, The method comprises the following steps: S1, collecting the current signal and the rotating speed signal of the shunt DC motor without stopping, carrying out band-pass filtering on the current signal to obtain high-frequency current filtered data, and carrying out low-frequency filtering on the rotating speed signal to obtain low-frequency rotating speed filtered data; S2, carrying out wavelet decomposition on the current filtered data to obtain real-time current detail signal data and real-time current approximation signal data, calculating the energy change rate of the real-time current detail signal data relative to the current detail signal data under normal operating state, and determining the spark deviation fault characteristic quantity and the current fluctuation amplitude based on the energy change rate and a preset fault diagnosis threshold value; S3, comprehensively judging whether the shunt DC motor has the spark deviation fault based on the change trend of the rotating speed filtered data and the current fluctuation amplitude; The current signal comprises an armature current signal and an excitation current signal.

2. The method of detecting spark blowout of a shunt DC motor according to claim 1, wherein The S1 step specifically comprises the following steps: S11, collecting the armature current signal and the excitation current signal of the shunt DC motor by using a Hall current sensor, and collecting the rotating speed signal of the shunt DC motor by using a non-contact signal acquisition system; S12, carrying out band-pass filtering on the armature current signal and the excitation current signal to obtain high-frequency armature current filtered data and excitation current filtered data, and carrying out low-frequency filtering on the rotating speed signal to obtain low-frequency rotating speed filtered data.

3. The method of claim 2, wherein the step of detecting the spark deviation is characterized by, The S2 step specifically comprises the following steps: S21, carrying out wavelet decomposition on the armature current filtered data and the excitation current filtered data by using a db8 wavelet function to obtain real-time armature current detail signal data, real-time armature current approximation signal data, real-time excitation current detail signal data and real-time excitation current approximation signal data composed of different frequency layers; S22, calculating the energy corresponding to each layer of the real-time armature current detail signal data and the real-time armature current approximate signal data, and obtaining a first energy change rate AW of the energy corresponding to the armature current detail signal data and the approximate signal data in the normal operating state j ; S23, calculating the energy corresponding to each layer of the real-time field current detail signal data and the real-time field current approximate signal data, and obtaining a second energy change rate AQ of the energy corresponding to the field current detail signal data and the approximate signal data in the normal operating state j ; S23, preset the fault diagnosis threshold of the armature current corresponding to the jth layer as AW j and the fault diagnosis threshold of the field current as AQ j , when AW j > AW j or AQ j > AQ j , the jth layer armature current data or the field current data is determined as the spark deviation fault characteristic quantity; Wherein, j is a positive integer, and j is less than or equal to the wavelet decomposition layer number.

4. The method of detecting spark blowout of a shunt DC machine according to claim 3, characterized by, the energy corresponding to each layer of the real-time armature current detail signal data The calculation formula is: the energy corresponding to each layer of the real-time excitation current detail signal data The calculation formula is: The first energy change rate ΔW j The calculation formula is: The second energy change rate AQ j The calculation formula is: wherein, is the energy corresponding to the jth layer of the armature current detail signal data in the normal operating state, is the energy corresponding to the jth layer of the field current detail signal data in the normal operating state, N j is the length of the jth layer of detail coefficients, k represents the kth data of the jth layer of wavelet decomposition detail signal, and d represents the detail signal.

5. The method of detecting spark blowout of a shunt DC machine according to claim 3, wherein The S23 step specifically comprises the following steps: the fault diagnosis threshold value AW of the armature current corresponding to the jth layer j is 40%, and the fault diagnosis threshold value AQ of the exciting current corresponding to the jth layer j is 40%.

6. The method of detecting spark blowout of a shunt DC machine according to claim 3, wherein The S3 step specifically comprises the following steps: S31, determining the energy corresponding to each layer of the real-time armature current detail signal data based on the energy corresponding to each layer of the real-time armature current detail signal data , the energy corresponding to each layer of the real-time field current detail signal data , the energy corresponding to each layer of the normal operation armature current detail signal data is , and the energy corresponding to each layer of the normal operation field current detail signal data is , determining and in relation to and the n-th layer with the largest comprehensive change of energy is the observation layer; S32, observing the up-and-down vibration of the wavelet in the observation layer within i cycles, counting the actual current fluctuation proportion of the number of times of up-and-down vibration exceeding the vibration threshold value within i cycles, and comparing the actual current fluctuation proportion with a preset safe fluctuation proportion to obtain the current fluctuation amplitude; S33, comprehensively considering the change state of the rotating speed filtered data and the current fluctuation amplitude, if any of the following conditions is met, it is judged that the shunt DC motor has the spark deviation fault, otherwise, it is judged that the shunt DC motor does not have the spark deviation fault: A1) the waveform change trend of the rotating speed filtered data is not gradually decreasing first and then tending to be stable and slightly oscillating; A2) the actual current fluctuation proportion is higher than the safe fluctuation proportion, that is, the current fluctuation amplitude is too large; Wherein, n and i are positive integers.

7. The method of detecting spark blowout of a shunt DC machine according to claim 6, wherein The S32 step specifically comprises the following steps: S321, observing the up-and-down vibration of the wavelet in the observation layer within i cycles, counting the number m of times of up-and-down vibration exceeding the vibration threshold value within i cycles, and calculating the actual current fluctuation proportion Z1; S322, presetting a safe fluctuation proportion Z0, comparing the actual current fluctuation proportion Z1 with the preset safe fluctuation proportion Z0 to obtain the current fluctuation amplitude; If Z1>Z0, the current fluctuation amplitude is too large, otherwise, the current fluctuation amplitude is normal; The actual current fluctuation ratio Z1=m / i, m is a positive integer.

8. A device for detecting excessive sparking in a shunt-wound DC motor, characterized in that, The shunt DC motor spark detection device comprises a collection box and an analysis system. The signal input end of the collection box is connected with the shunt DC motor to be measured, for collecting the current signal and the speed signal of the shunt DC motor without stopping, and filtering the current signal and the speed signal respectively to obtain current filtered data and speed filtered data respectively. The signal input end of the analysis system is connected with the signal output end of the collection box, for receiving and analyzing the current filtered data and the speed filtered data from the collection box, and comprehensively judging whether the shunt DC motor has a spark fault. The shunt DC motor spark detection device is used to realize the shunt DC motor spark detection method as claimed in any one of claims 1-7.

9. The device for detecting spark over of a shunt DC machine according to claim 8, characterized by The collection box comprises a current collection module and a speed collection module. The current collection module is connected with the shunt DC motor to be measured, for collecting the current signal of the shunt DC motor and performing high-frequency filtering on the current signal to obtain high-frequency current filtered data. The speed collection module is connected with the shunt DC motor to be measured, for collecting the speed signal of the shunt DC motor and performing low-frequency filtering on the speed signal to obtain low-frequency speed filtered data.

10. The device for detecting spark over of a shunt DC machine according to claim 9, wherein The analysis system comprises a current signal conditioning module, a current signal analysis module, a speed signal analysis module and a comprehensive analysis module. The input end of the current signal conditioning module is connected with the output end of the current collection module, for receiving the current filtered data and performing wavelet decomposition to obtain real-time current detail signal data. The input end of the current signal analysis module is connected with the output end of the current signal conditioning module, for calculating the energy change rate of the real-time current detail signal data relative to the current detail signal data under normal operating state, determining the spark fault characteristic quantity and the current fluctuation amplitude based on the energy change rate and a preset fault diagnosis threshold. The input end of the speed signal analysis module is connected with the output end of the speed collection module, for receiving the speed filtered data and judging the waveform change state of the speed filtered data. The input end of the comprehensive analysis module is connected with the output end of the current signal analysis module and the output end of the speed signal analysis module, for integrating the waveform change state of the speed filtered data and the current fluctuation amplitude to judge whether the shunt DC motor has a spark fault.

Images