Electromechanical equipment fault diagnosis method and system

Abstract

The invention discloses an electromechanical equipment fault diagnosis method and system, relates to the technical field of fault diagnosis, and aims to solve the technical problem of improving the fault diagnosis capability of electromechanical equipment. The electromechanical equipment fault diagnosis method and system can obtain the running state data information of the electromechanical equipment and classify the collected data information; the operation state of the electromechanical equipment is monitored in a 24 uninterrupted mode, and output data information is amplified, so that analysis of the data information is realized; the method comprises the following steps: constructing an electromechanical equipment fault diagnosis mathematical model, adaptively adjusting data information through a fault early warning algorithm of particle filter state estimation, and analyzing electromechanical equipment fault information; and outputting the fault diagnosis data information of the electromechanical equipment so as to carry out remote online monitoring on the fault data information of the electromechanical equipment in real time. The fault diagnosis capability of the electromechanical equipment is greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of data diagnosis, and more specifically to a method and system for diagnosing mechanical and electrical equipment faults. Background Art

[0002] Electromechanical equipment generally refers to machinery, electrical appliances and electrical automation equipment. In construction, it refers to the general term for machinery and pipeline equipment other than geotechnical, woodworking, steel bars and mud and water equipment. It is different from hardware and mostly refers to finished products that can achieve certain functions. Advanced electromechanical equipment can not only greatly improve labor productivity, reduce labor intensity and improve the production environment. However, during the operation of electromechanical equipment, it is easy to have various failure factors, such as electromechanical equipment power failure and component failure, electromechanical equipment operation failure, parameter data information failure and other problems.

[0003] How to realize electromechanical equipment fault diagnosis is related to whether the electromechanical equipment can operate normally. The conventional method uses tools such as multimeters, oscilloscopes, and detection equipment to diagnose electromechanical equipment faults. Although this method has a certain intuitiveness, the electromechanical equipment fault diagnosis of this method cannot realize the microscopic analysis of data information, and the existing technology has poor analysis efficiency. Summary of the invention

[0004] In view of the deficiencies of the above-mentioned technologies, the present invention discloses a method and system for diagnosing faults of electromechanical equipment, which greatly improves the diagnostic and analytical capabilities of the faults of electromechanical equipment.

[0005] The present invention adopts the following technical solutions:

[0006] A method for diagnosing a fault in an electromechanical device, comprising the following method:

[0007] Step 1: Obtain the operating status data of the electromechanical equipment and classify the collected data information; Step 2: Monitor the operating status of the electromechanical equipment in a 24-hour uninterrupted manner, amplify the output data information, and then analyze the data information;

[0008] Step 3: Construct a mathematical model for mechanical and electrical equipment fault diagnosis, adaptively adjust data information through a fault warning algorithm based on particle filter state estimation, and analyze the mechanical and electrical equipment fault information;

[0009] Step 4: Output the electromechanical equipment fault diagnosis data information to remotely monitor the electromechanical equipment fault data information in real time.

[0010] As a further technical solution of the present invention, the working method of the fault warning algorithm of particle filter state estimation is:

[0011] Realize the correlation calculation of fault warning data information through uncertainty matrix theory;

[0012] The correlation calculation of fault warning data information is realized through particle filtering algorithm.

[0013] As a further technical solution of the present invention, the method of deterministic matrix theory is:

[0014] Constructing the mathematical model of uncertainty matrix theory:

[0015]

[0016] In formula (1), C represents the total matrix of the mathematical model of uncertainty matrix theory, C 1 C 2 The figures all represent some matrices in the uncertainty matrix theory model, which are specifically:

[0017]

[0018] Among them, C 1 {D 1 ,D 2 ,D 3 ...D J} indicates that the elements that affect electromechanical equipment constitute a data set, D indicates one of the elements of the data set, the subscript is the serial number, and there are J elements of electromechanical equipment, namely magnetic field, operating environment and weather, and N indicates the number of data;

[0019] C 2 {L 1 ,L 2 ,L 3 ...L K} represents the data set composed of the elements that affect the electromechanical equipment, K represents one of the elements, and it is assumed that there are K kinds of elements of the electromechanical equipment;

[0020] When looking for factors affecting electromechanical equipment, connect the two sets, observe the similarities or differences between them, and determine the elements that affect the electromechanical equipment.

[0021] As a further technical solution of the present invention, the working method of the particle filter algorithm is:

[0022] Assume that the probability density function of the electromechanical equipment fault status data information at time t-1 is P(x t-1 |y t-1 ), sample N values ​​x t-1 (i) , weight is w t-1 (i) The particles are used to estimate the prediction probability density P(x t |y t-1 ), the new particle is expressed as:

[0023]

[0024] In formula (4), x t-1 (i) represents the value of the i-th particle at time t-1, w t-1 (i) represents the weight of the i-th particle at time t-1. In the state update phase, the state at time t is predicted based on the historical observations. t After that, the weight of the particle is determined, which is expressed as:

[0025]

[0026] In formula (5), P(y t |x (i) t|t-1 ) represents the likelihood function. In the resampling stage, the particles are resampled according to their weights, particles with larger weights are retained, particles with smaller weights are removed, and the particle degradation problem is reduced. The particle weight after resampling is 1 / N. The estimation of the current state variable is expressed as:

[0027]

[0028] In formula (6), Z (i) t Represents the value of the particle at time t; after the particle filtering process, the particle filter state estimator of the electromechanical equipment is realized, and the difference between the estimated value and the actual value of the state variable is used to warn the equipment failure. The variable residual is expressed as:

[0029]

[0030] In formula (7), y (l) t represents the actual measured value of the state variable, x (l) t Represents the particle filter estimation value, and uses the multi-state variable residual as a fault warning indicator, expressed as:

[0031]

[0032] In formula (8), r (l) k Represents the state quantity residual;

[0033] When the equipment is operating normally, some state variables will be affected by load changes, and the amplitude of the vibration signal will be affected by the load current on the equipment conductor. Therefore, the state variable residual may not be 0. In order to minimize errors and noise interference,

[0034] When the fault occurs at t AAt time t, if the constant threshold method is used, B The alarm signal is always issued, and false alarm occurs. The use of the adaptive threshold method can effectively avoid other interference factors affecting the change of variable residuals, thereby making the equipment fault alarm function more accurate. The confidence and confidence level of the adaptive threshold satisfy the following formula:

[0035]

[0036] In formula (9), α represents the confidence level, 1-α represents the confidence, and Z represents the correlation coefficient.

[0037] A mechanical and electrical equipment fault diagnosis system, comprising:

[0038] The data acquisition module acquires the operating status data information of the electromechanical equipment and classifies the acquired data information;

[0039] The monitoring module monitors the operating status of electromechanical equipment 24 hours a day, amplifies the output data information, and then analyzes the data information;

[0040] A diagnosis module, which is a mathematical model for diagnosing electromechanical equipment faults, adaptively adjusts data information through a fault warning algorithm based on particle filter state estimation, and analyzes the electromechanical equipment fault information;

[0041] Output module, outputs electromechanical equipment fault diagnosis data information, so as to remotely monitor the electromechanical equipment fault data information in real time;

[0042] The output end of the data acquisition module is connected to the input end of the monitoring module, the output end of the monitoring module is connected to the input end of the diagnosis module, and the output end of the diagnosis module is connected to the input end of the output module.

[0043] As a further technical solution of the present invention, the data acquisition module includes a preamplifier circuit based on an INA163 chip.

[0044] As a further technical solution of the present invention, the monitoring module adopts the STM32F103RCT6 single chip microcomputer to realize the monitoring of circuit data information.

[0045] As a further technical solution of the present invention, the diagnostic module is provided with a data conversion interface, and the data conversion interface is used to convert communication data information into artificial intelligence programmable data information.

[0046] As a further technical solution of the present invention, the output module is provided with a visual display screen, and the output of data information is realized through the visual display screen.

[0047] The present invention has the following positive and beneficial effects:

[0048] The present invention can obtain the data information of the running status of electromechanical equipment, and classify the collected data information; monitor the running status of electromechanical equipment in a 24-hour uninterrupted manner, amplify the output data information, and then realize the analysis of the data information; construct a mathematical model for fault diagnosis of electromechanical equipment, adaptively adjust the data information through the fault warning algorithm of particle filter state estimation, and analyze the fault information of electromechanical equipment; output the fault diagnosis data information of electromechanical equipment, so as to remotely monitor the fault data information of electromechanical equipment in real time. The present invention greatly improves the fault diagnosis capability of electromechanical equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work, among which:

[0050] Figure 1 It is the overall architecture of the differential photoacoustic spectroscopy gas detection system;

[0051] Figure 2 It is a schematic diagram of the state monitoring circuit of the present invention;

[0052] Figure 3 This is a schematic diagram of the adaptive threshold adjustment of the present invention;

[0053] Figure 4 This is a schematic diagram of the preamplifier sampling circuit. DETAILED DESCRIPTION

[0054] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0055] like Figure 1-Figure 4 As shown, a method for diagnosing a fault of an electromechanical device includes the following methods:

[0056] Step 1: Obtain the operating status data of the electromechanical equipment and classify the collected data information; Step 2: Monitor the operating status of the electromechanical equipment in a 24-hour uninterrupted manner, amplify the output data information, and then analyze the data information;

[0057] Step 3: Construct a mathematical model for mechanical and electrical equipment fault diagnosis, adaptively adjust data information through a fault warning algorithm based on particle filter state estimation, and analyze the mechanical and electrical equipment fault information;

[0058] Step 4: Output the electromechanical equipment fault diagnosis data information to remotely monitor the electromechanical equipment fault data information in real time.

[0059] In the above embodiment, the working method of the fault warning algorithm of particle filter state estimation is:

[0060] Realize the correlation calculation of fault warning data information through uncertainty matrix theory;

[0061] The correlation calculation of fault warning data information is realized through particle filtering algorithm.

[0062] In the above embodiment, the method of deterministic matrix theory is:

[0063] Constructing the mathematical model of uncertainty matrix theory:

[0064]

[0065] In formula (1), C represents the total matrix of the mathematical model of uncertainty matrix theory, C 1 C 2 The figures all represent some matrices in the uncertainty matrix theory model, which are specifically:

[0066]

[0067] Among them, C 1 {D 1 ,D 2 ,D 3 ...D J} indicates that the elements that affect electromechanical equipment constitute a data set, D indicates one of the elements of the data set, the subscript is the serial number, and there are J elements of electromechanical equipment, namely magnetic field, operating environment and weather, and N indicates the number of data;

[0068] C 2 {L 1 ,L 2 ,L 3 ...L K} represents the data set composed of the elements that affect the electromechanical equipment, K represents one of the elements, and it is assumed that there are K kinds of elements of the electromechanical equipment;

[0069] When looking for factors affecting electromechanical equipment, connect the two sets, observe the similarities or differences between them, and determine the elements that affect the electromechanical equipment.

[0070] In order to reflect the nonlinear characteristics of the state characteristic variables of electromechanical equipment and their characteristics that change with the seasons, it is possible to effectively warn of equipment failures. The traditional Kalman filter method can only process linear systems, while the present invention uses a filtering algorithm to directly estimate the state variables of electromechanical equipment, which expands the Kalman filter method and enables the system to perform linear state estimation on nonlinear systems. Combining the importance sampling method and the particle filter algorithm, the process of estimating the state variables of electromechanical equipment is divided into three stages: state prediction, state update, and resampling.

[0071] In the above embodiment, the working method of the particle filter algorithm is:

[0072] Assume that the probability density function of the electromechanical equipment fault status data information at time t-1 is P(x t-1 |y t-1 ), sample N values ​​x t-1 (i) , weight is w t-1 (i) The particles are used to estimate the prediction probability density P(x t |y t-1 ), the new particle is expressed as:

[0073]

[0074] In formula (4), x t-1 (i) represents the value of the i-th particle at time t-1, w t-1 (i) represents the weight of the i-th particle at time t-1. In the state update phase, the state at time t is predicted based on the historical observations. t After that, the weight of the particle is determined, which is expressed as:

[0075]

[0076] In formula (5), P(y t |x (i) t|t-1 ) represents the likelihood function. In the resampling stage, the particles are resampled according to their weights, particles with larger weights are retained, particles with smaller weights are removed, and the particle degradation problem is reduced. The particle weight after resampling is 1 / N. The estimation of the current state variable is expressed as:

[0077]

[0078] In formula (6), Z (i) tRepresents the value of the particle at time t; after the particle filtering process, the particle filter state estimator of the electromechanical equipment is realized, and the difference between the estimated value and the actual value of the state variable is used to warn the equipment failure. The variable residual is expressed as:

[0079]

[0080] In formula (7), y (l) t represents the actual measured value of the state variable, x (l) t Represents the particle filter estimation value, and uses the multi-state variable residual as a fault warning indicator, expressed as:

[0081]

[0082] In formula (8), r (l) k Represents the state quantity residual;

[0083] When the equipment is operating normally, some state variables will be affected by load changes, and the amplitude of the vibration signal will be affected by the load current on the equipment conductor, so the state variable residual may not be 0. In order to minimize the error and noise interference, the present invention introduces an adaptive threshold method to detect equipment failures;

[0084] When the fault occurs at t A At time t, if the constant threshold method is used, B The alarm signal is always issued, and false alarm occurs. The use of the adaptive threshold method can effectively avoid other interference factors affecting the change of variable residuals, thereby making the equipment fault alarm function more accurate. The confidence and confidence level of the adaptive threshold satisfy the following formula:

[0085]

[0086] In formula (9), α represents the confidence level, 1-α represents the confidence, and Z represents the correlation coefficient.

[0087] The state variable data of the electromechanical equipment during normal operation is used, the particle calculation filter algorithm is adopted, and the state estimator is established with optimized parameters and initial conditions. By comparing the current actual measurement value and the estimated value of the state variable, the variable residual is calculated to obtain the state warning index. Combined with the adaptive threshold, the detection of electromechanical equipment failure is realized. When a variable residual of the equipment state exceeds the adaptive threshold, it indicates that the equipment has a fault and the model issues an alarm.

[0088] A mechanical and electrical equipment fault diagnosis system, comprising:

[0089] The data acquisition module acquires the operating status data information of the electromechanical equipment and classifies the acquired data information;

[0090] The monitoring module monitors the operating status of electromechanical equipment 24 hours a day, amplifies the output data information, and then analyzes the data information;

[0091] A diagnosis module, which is a mathematical model for diagnosing electromechanical equipment faults, adaptively adjusts data information through a fault warning algorithm based on particle filter state estimation, and analyzes the electromechanical equipment fault information;

[0092] The output module outputs the electromechanical equipment fault diagnosis data information to remotely monitor the electromechanical equipment fault data information in real time.

[0093] The output end of the data acquisition module is connected to the input end of the monitoring module, the output end of the monitoring module is connected to the input end of the diagnosis module, and the output end of the diagnosis module is connected to the input end of the output module.

[0094] The data acquisition module includes a preamplifier circuit based on the INA163 chip.

[0095] In the preamplifier circuit, in order to amplify the weak signal to a suitable voltage for sampling, the present invention sets a preamplifier circuit based on the INA163 chip. The present invention uses the AS-B4015AL38RC-L50 microphone, which has a sensitivity of -35dB, an impedance of 2.2KΩ, and a frequency response of 16 to 20KHz. The microphone pin adopts the SMA port to connect the peripheral bias circuit, adopts a 3.3V DC power supply, and a 0.22uF capacitor is connected in series in the signal channel to block the DC signal. The collected photoacoustic signal is passed through a signal processing circuit with high gain and noise reduction function, amplified and noise-reduced, and then transmitted to the DSP chip for sampling processing. The amplification factor of the preamplifier circuit can be controlled and adjusted by the RH resistor, and the amplified signal enters the A / D conversion port of the PCM1804, and then the converted information is transmitted to the ADSP21262 chip for processing. The INA162 amplifier is a device that integrates four high-precision resistor pairs and on-chip EMI filtering with a low-distortion, high-output current, dual-channel audio operational amplifier. It has a strong suppression capability for circuit noise. The main chip operates at 200MHz, the 1024-point complex FFT execution time is 46uS, and it has good dynamic response.

[0096] In the above embodiment, the monitoring module uses the STM32F103RCT6 single chip microcomputer to monitor the circuit data information.

[0097] When monitoring the circuit, the monitoring circuit uses the STM32F103RCT6 single-chip microcomputer, and the independent wireless alarm uses the STM8L151K4T6 single-chip microcomputer to detect the circuit data information. Since the single-chip microcomputer of the field centralized controller has a low-power standby function, the standby current is only 1 to 2 microamperes when maintaining its minimum system operation. Considering the standby power consumption problem and circuit simplification, the STM32F103RCT6 single-chip microcomputer is used as the core processing chip. The maximum clock frequency is 72MHz, the RAM size is 32768B, the FLASH memory capacity is 262144B, the number of pins is 64, it has 51 high-speed input and output ports, the operating voltage is DC2V~DC3.6V, and the operating temperature is -40℃-85℃.

[0098] The signal conditioning circuit no longer uses the XD1129 hub chip, but instead transforms the chip part into a combination of a bridge rectifier circuit and a triode, which becomes a combination of two circuit parts, adds an inductor, and adds an electrolytic capacitor with a larger capacitance, which can also play a filtering role and stabilize the current. The working principle is basically the same as the above figure. The analog signal still enters the rectifier circuit, passes through the triode, and then outputs to the AD channel in the MCU. The advantage of this circuit is that the voltage can be made more stable through bridge rectification, and then after passing through the triode, the signal can be amplified into a larger amplitude electrical signal, so that the signal changes can be monitored more accurately, and the accuracy of monitoring electromechanical equipment can be improved.

[0099] The diagnostic module is provided with a data conversion interface, and the data conversion interface is used to convert communication data information into artificial intelligence programmable data information.

[0100] In a specific embodiment, the diagnosis module can realize the conversion of data information in real time by setting a data conversion interface, and can realize the conversion between different data parameters in the data programmable program.

[0101] In a specific embodiment, the output module is provided with a visual display screen, and the output of data information is realized through the visual display screen.

[0102] In a specific embodiment, the visual display screen can be connected to the wireless data interface to achieve visual monitoring of different data information.

[0103] Although the specific embodiments of the present invention are described above, it should be understood by those skilled in the art that these specific embodiments are only illustrative, and those skilled in the art may omit, replace, and change the details of the above methods and inventions in various ways without departing from the principles and essence of the present invention. For example, merging the above method steps so as to perform substantially the same functions in substantially the same manner to achieve substantially the same results is within the scope of the present invention. Therefore, the scope of the present invention is limited only by the appended claims.

Claims

1. A method for diagnosing mechanical and electrical equipment faults, characterized in that: The following methods are included: Step 1: Obtain the operating status data information of the electromechanical equipment and classify the collected data information; Step 2: Monitor the operating status of electromechanical equipment 24 hours a day, amplify the output data information, and then analyze the data information; Step 3: Construct a mathematical model for mechanical and electrical equipment fault diagnosis, adaptively adjust data information through a fault warning algorithm based on particle filter state estimation, and analyze the mechanical and electrical equipment fault information; Step 4: Output the electromechanical equipment fault diagnosis data information to remotely monitor the electromechanical equipment fault data information in real time.

2. A method for diagnosing mechanical and electrical equipment faults according to claim 1, characterized in that: The working method of the fault warning algorithm of particle filter state estimation is: Realize the correlation calculation of fault warning data information through uncertainty matrix theory; The correlation calculation of fault warning data information is realized through particle filtering algorithm.

3. A method for diagnosing mechanical and electrical equipment faults according to claim 1, characterized in that: The method of deterministic matrix theory is: Constructing the mathematical model of uncertainty matrix theory: In formula (1), C represents the total matrix of the mathematical model of uncertainty matrix theory, and C1 and C2 represent partial matrices in the uncertainty matrix theory model, which are specifically: Among them, {D1,D2,D3...D J } indicates that the elements that affect electromechanical equipment constitute a data set, D indicates one of the elements of the data set, the subscript is the serial number, and there are J elements of electromechanical equipment, namely magnetic field, operating environment and weather, and N indicates the number of data; {L1,L2,L3...L K } represents the data set composed of elements that affect electromechanical equipment, K represents one of the elements, and there are K elements of electromechanical equipment; When looking for factors affecting electromechanical equipment, connect the two sets, observe the similarities or differences between them, and determine the elements that affect the electromechanical equipment.

4. A method for diagnosing mechanical and electrical equipment faults according to claim 1, characterized in that: The working method of the particle filter algorithm is: Assume that the probability density function of the electromechanical equipment fault status data information at time t-1 is P(x t-1 |y t-1 ), sample N values ​​x t-1 (i) , weight is w t-1 (i) The particles are used to estimate the prediction probability density P(x t |y t-1 ), the new particle is expressed as: In formula (4), x t-1 (i) represents the value of the i-th particle at time t-1, w t-1 (i) represents the weight of the i-th particle at time t-1. In the state update phase, the state at time t is predicted based on the historical observations. t After that, the weight of the particle is determined, which is expressed as: In formula (5), P(y t |x (i) t|t-1 ) represents the likelihood function. In the resampling stage, the particles are resampled according to their weights, particles with larger weights are retained, particles with smaller weights are removed, and the particle degradation problem is reduced. The particle weight after resampling is 1 / N. The estimation of the current state variable is expressed as: In formula (6), Z (i) t Represents the value of the particle at time t; after the particle filtering process, the particle filter state estimator of the electromechanical equipment is realized, and the difference between the estimated value and the actual value of the state variable is used to warn the equipment failure. The variable residual is expressed as: In formula (7), y (l) t represents the actual measured value of the state variable, x (l) t Represents the particle filter estimation value, and uses the multi-state variable residual as a fault warning indicator, expressed as: In formula (8), r (l) k Represents the state quantity residual; When the equipment is operating normally, some state variables will be affected by load changes, and the amplitude of the vibration signal will be affected by the load current on the equipment conductor. Therefore, the state variable residual may not be 0. In order to minimize errors and noise interference, When the fault occurs at t A At time t, if the constant threshold method is used, B The alarm signal is always issued, and false alarm occurs. The use of the adaptive threshold method can effectively avoid other interference factors affecting the change of variable residuals, thereby making the equipment fault alarm function more accurate. The confidence and confidence level of the adaptive threshold satisfy the following formula: In formula (9), α represents the confidence level, 1-α represents the confidence, and Z represents the correlation coefficient.

5. A mechanical and electrical equipment fault diagnosis system, characterized in that: include: The data acquisition module acquires the operating status data information of the electromechanical equipment and classifies the acquired data information; The monitoring module monitors the operating status of electromechanical equipment 24 hours a day, amplifies the output data information, and then analyzes the data information; A diagnosis module, which is a mathematical model for diagnosing electromechanical equipment faults, adaptively adjusts data information through a fault warning algorithm based on particle filter state estimation, and analyzes the electromechanical equipment fault information; Output module, outputs electromechanical equipment fault diagnosis data information, so as to remotely monitor the electromechanical equipment fault data information in real time; The output end of the data acquisition module is connected to the input end of the monitoring module, the output end of the monitoring module is connected to the input end of the diagnosis module, and the output end of the diagnosis module is connected to the input end of the output module.

6. A mechanical and electrical equipment fault diagnosis system according to claim 5, characterized in that: The data acquisition module includes a preamplifier circuit based on an INA163 chip.

7. The electromechanical equipment fault diagnosis system according to claim 5, characterized in that: The monitoring module uses the STM32F103RCT6 microcontroller to monitor circuit data information.

8. The electromechanical equipment fault diagnosis system according to claim 5, characterized in that: The diagnostic module is provided with a data conversion interface, and the data conversion interface is used to convert communication data information into artificial intelligence programmable data information.

9. The electromechanical equipment fault diagnosis system according to claim 5, characterized in that: The output module is provided with a visual display screen, through which the data information is output.

Images