CN118258606A - Bearing fault diagnosis method for traction motor of electric forklift in complex noise environment - Google Patents
Bearing fault diagnosis method for traction motor of electric forklift in complex noise environment Download PDFInfo
- Publication number
- CN118258606A CN118258606A CN202410395053.1A CN202410395053A CN118258606A CN 118258606 A CN118258606 A CN 118258606A CN 202410395053 A CN202410395053 A CN 202410395053A CN 118258606 A CN118258606 A CN 118258606A
- Authority
- CN
- China
- Prior art keywords
- fault
- kth
- traction motor
- unit
- phase
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 37
- 238000003745 diagnosis Methods 0.000 title claims abstract description 26
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 21
- 238000012549 training Methods 0.000 claims abstract description 12
- 238000012545 processing Methods 0.000 claims abstract description 9
- 230000006835 compression Effects 0.000 claims abstract description 7
- 238000007906 compression Methods 0.000 claims abstract description 7
- 230000006870 function Effects 0.000 claims description 13
- 238000012937 correction Methods 0.000 claims description 6
- 230000002159 abnormal effect Effects 0.000 claims description 5
- 238000004590 computer program Methods 0.000 claims description 5
- 238000011478 gradient descent method Methods 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 238000003860 storage Methods 0.000 claims description 3
- 238000004148 unit process Methods 0.000 claims description 3
- 238000005259 measurement Methods 0.000 claims description 2
- 238000001514 detection method Methods 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 3
- 230000032683 aging Effects 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001364 causal effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/04—Bearings
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R19/00—Arrangements for measuring currents or voltages or for indicating presence or sign thereof
- G01R19/0092—Arrangements for measuring currents or voltages or for indicating presence or sign thereof measuring current only
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/34—Testing dynamo-electric machines
- G01R31/343—Testing dynamo-electric machines in operation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computational Linguistics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Control Of Electric Motors In General (AREA)
Abstract
The invention discloses a bearing fault diagnosis method of an electric forklift traction motor in a complex noise environment, which comprises the following steps: 1. collecting working phase current of an electric fork-lift motor under a normal working environment; 2. constructing a phase current data set and constructing a one-dimensional convolutional neural network, comprising: each hidden layer comprises a one-dimensional convolution unit, a compression unit, a linear rectification unit and a regularization unit; 3. the network processes the data and obtains a fault prediction result; 4. and training and optimizing the convolutional neural network to obtain a trained electric forklift traction motor bearing fault diagnosis model. The invention has better environment noise anti-interference capability, has high accuracy on data acquisition and processing, can timely and accurately reflect the mechanical condition of the traction motor bearing of the electric forklift when the motor normally operates, and has extremely high practical value in engineering practical application.
Description
Technical Field
The invention belongs to the field of bearing fault diagnosis of common motors, and particularly relates to a bearing fault diagnosis method of an electric forklift traction motor in a complex noise environment.
Background
Traction motors are the primary power means for driving the vehicle concerned, the function of which can be summarized as converting electrical energy in a battery pack into mechanical energy for driving the vehicle. Electric forklifts are now widely used in a variety of logistic transportation and industrial production activities, which require reliable, high quality traction motors to maintain their proper operation.
Through investigation, bearing failure is one of the most common failures in the everyday use of electric forklift traction motors. The motor bearing faults are often caused by slow aging of the bearing itself to reach the fatigue limit or slight damage to enlarge cracks after severe vibration, and the slow aging or the slight damage is difficult to monitor and find in normal operation; in addition, the working environment of the forklift motor belongs to a complex noise environment, and for general vibration signal detection, the fault detection result is inaccurate and frequently happens due to excessive interference data. Therefore, it is an extremely complex task to detect bearing faults in a timely and accurate manner by the existing fault detection techniques.
In addition, the neural network commonly used at present can cause conditions of low precision or time consumption and the like in the process of processing related data sequences in the fault detection and analysis of the traction motor of the electric forklift.
Disclosure of Invention
The invention aims to solve the defects of the prior art, and provides a bearing fault diagnosis method for an electric forklift traction motor under a complex noise environment, so that a current signal during fault can be accurately analyzed, whether bearing faults occur and which type of bearing faults belong to can be accurately judged, and the timeliness and the accuracy of fault diagnosis are improved.
In order to achieve the aim of the invention, the invention adopts the following technical scheme:
the invention relates to a bearing fault diagnosis method for an electric forklift traction motor in a complex noise environment, which is characterized by comprising the following steps:
step 1: acquiring a single-phase fault current data set I t={i1,t,i2,t,i3,t,...,in,t,...,iN,t of any one of phase A current, phase B current and phase C current of a traction motor of the electric forklift at the time T by using a phase current measurement link, wherein I n,t represents the nth single-phase current at the time T, and T is more than or equal to 0 and less than or equal to T s,Ts and is the total sampling time; n is more than 0 and less than or equal to N, wherein N is the total number of fault phase currents;
Let i n,t have a tag value of b n,t, where b e [1, D ], D is the number of failed species;
step 2: constructing a one-dimensional convolutional neural network, comprising: an input layer, K hidden layers and an output layer; wherein each hidden layer comprises: the device comprises a one-dimensional convolution unit, a compression unit, a linear rectification unit and a regularization unit;
Step 2.1: the input layer carries out convolution processing on the input i n,t to obtain a fault convolution characteristic i n,t *;
Step 2.2: when k=1, i n,t * is input into a kth hidden layer, a one-dimensional convolution unit processes i n,t * to obtain a kth information hidden state W n,t,k, a compression unit of the kth hidden layer processes the kth information hidden state W n,t,k to obtain a kth fault information feature X n,t,k, a linear rectification unit of the kth hidden layer processes X n,t,k to obtain a kth fault correction feature z n,t,k,zn,t,k, and the kth fault correction feature z n,t,k,zn,t,k is input into a regularization unit of the k hidden layer to process the kth fault information feature J n,t,k;
When k=2, 3, … and K, J n,t,k-1 inputs the K hidden layer for processing to obtain a K fault feature J n,t,k, and finally the K hidden layer outputs a K fault feature J n,t,K;
Step 2.3: the output layer processes J n,t,K to obtain a predicted tag value b n,t * of i n,t and outputs the predicted tag value b n,t *;
Step 4: based on b n,t and b n,t *, constructing a cross entropy loss function of a one-dimensional convolutional neural network, calculating the cross entropy loss function to update network parameters when training the convolutional neural network by using a gradient descent method, and stopping training until the cross entropy loss function converges, so as to obtain a trained electric fork truck traction motor bearing fault diagnosis model, and performing fault prediction on abnormal phase current when the electric fork truck traction motor works.
The electronic device of the present invention includes a memory for storing a program for supporting the processor to execute the bearing failure diagnosis method, and a processor configured to execute the program stored in the memory.
The invention relates to a computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being run by a processor, performs the steps of the bearing fault diagnosis method.
Compared with the prior art, the invention has the beneficial effects that:
1. According to the invention, whether the bearing of the traction motor of the electric forklift is faulty or not is indirectly analyzed by analyzing the change or abnormality of the current signal, and monitoring equipment is not required to be directly arranged at the bearing of the traction motor of the electric forklift, so that the complexity and inaccuracy of directly detecting the fault of the bearing are avoided, and the service life of corresponding diagnosis equipment can be prolonged by adopting the fault diagnosis mode;
2. According to the invention, the diagnosis of the bearing faults of the traction motor of the electric forklift can be completed by analyzing the current signals, the coordination with other fault diagnosis models can be realized, other faults of the motor can also cause abnormal changes of certain current signals, and the abnormal changes are different from the current signal law of the bearing faults, so that the model can be upgraded in the future to diagnose various common fault forms of the traction motor of the electric forklift, including the bearing faults, as long as the fault types are strictly distinguished in the model;
3. compared with a two-dimensional convolutional neural network, the method has the advantages that the one-dimensional convolutional neural network is adopted to process the current data set, the network structure is simpler, the efficiency is higher, and the accuracy is guaranteed;
4. The invention trains the bearing fault diagnosis model of the electric forklift traction motor by utilizing the convergence of the cross entropy loss function, and the specific process comprises the steps of calculating the cross entropy loss function to update network parameters, and stopping training until the cross entropy loss function converges, so that the trained electric forklift traction motor bearing fault diagnosis model is obtained, the accuracy of fault diagnosis results is improved, the interference received under the complex noise environment is small, and the bearing fault of the electric forklift traction motor can be diagnosed in time.
Drawings
FIG. 1 is a schematic flow chart of a method for diagnosing bearing faults of an electric fork-lift traction motor in a complex noise environment;
FIG. 2 is a training flow chart of a one-dimensional convolutional neural network model provided by the invention;
Fig. 3 is a schematic diagram of a one-dimensional convolutional neural network according to the present invention.
Detailed Description
In this embodiment, a bearing fault diagnosis method for an electric forklift traction motor in a complex noise environment is based on different current signal waveform change rules caused by various bearing faults of the electric forklift traction motor, and compares a current signal during working with a normal signal to discriminate whether the bearing fault occurs in the electric forklift traction motor. As shown in fig. 1, the bearing fault diagnosis method includes the steps of:
Firstly, measuring line current by using a phase current measuring link, and further obtaining a single-phase fault current data set I t={i1,t,i2,t,i3,t,...,in,t,...,iN,t of any one of phase current of an A phase, a B phase and a C phase of a traction motor of an electric forklift at a time T, wherein I n,t represents an nth single-phase current at the time T, and T is more than or equal to 0 and less than or equal to T s,Ts and is the total sampling time; n is more than 0 and less than or equal to N, N is the total number of fault phase currents, and then, the label value of i n,t is b n,t, wherein b epsilon [1, D ], and D is the number of fault types.
Then, a one-dimensional convolutional neural network is constructed, comprising: an input layer, K hidden layers and an output layer; wherein each hidden layer comprises: the device comprises a one-dimensional convolution unit, a compression unit, a linear rectification unit and a regularization unit.
As shown in fig. 2, a model training flowchart of the one-dimensional convolutional neural network used in the present embodiment is shown. Inputting the current data set into a one-dimensional convolutional neural network to be built, setting a network structure environment, configuring parameters, transmitting signals forward into the network, processing initial signals, comparing the initial signals with standard values, calculating errors, smoothly outputting the errors when the requirements are met, resetting the parameters when the requirements are not met, and repeating the processes until all current signals can be smoothly output, namely marking that the model training of the one-dimensional convolutional neural network is finished.
As shown in fig. 3, a schematic structure of the one-dimensional convolutional neural network is shown.
Inputting the current signal data set into an input layer, and carrying out convolution processing on the input i n,t by the input layer to obtain a fault convolution characteristic i n,t *; then, when k=1, i n,t * is input into a kth hidden layer, a one-dimensional convolution unit processes i n,t * to obtain a kth information hidden state W n,t,k, a compression unit of the kth hidden layer processes the kth information hidden state W n,t,k to obtain a kth fault information feature X n,t,k, a linear rectification unit of the kth hidden layer processes X n,t,k to obtain a kth fault correction feature z n,t,k,zn,t,k, and the kth fault correction feature z n,t,k,zn,t,k is input into a regularization unit of the k hidden layer to process the kth fault information feature J n,t,k;
Similarly, when k=2, 3, … and K, J n,t,k-1 inputs the K hidden layer for processing to obtain a K fault feature J n,t,k, and finally the K hidden layer outputs a K fault feature J n,t,K; finally, the output layer processes J n,t,K to obtain the predicted tag value b n,t * of i n,t, and outputs the predicted tag value b n,t *.
Next, based on b n,t and b n,t *, constructing a cross entropy loss function of a one-dimensional convolutional neural network, calculating the cross entropy loss function to update network parameters when training the convolutional neural network by using a gradient descent method, and stopping training until the cross entropy loss function converges, thereby obtaining a trained electric forklift traction motor bearing fault diagnosis model for carrying out fault prediction on abnormal phase current when the electric forklift traction motor works.
In this embodiment, an electronic device includes a memory for storing a program supporting the processor to execute the above method, and a processor configured to execute the program stored in the memory.
In this embodiment, a computer-readable storage medium stores a computer program that, when executed by a processor, performs the steps of the method described above.
In addition, the neural network of the present invention has made significant progress in both training time and accuracy, and network depth is enhanced to increase the potential for acceptance domains and overall accuracy.
One-dimensional convolutional neural networks not only have many advantages over two-dimensional convolutional neural networks, but also exhibit superior performance over RNNs. Moreover, the causal relationship of the use of the method relieves the general limitations (gradient explosion and insufficient memory) of the circulation model. The neural network used in the invention enables the network to rely on front and back by removing links layer by layer. At the same time, the network adopts a more reliable one-dimensional expansion convolution to expand the receptive field.
Compared with the traditional method, the method has excellent reliability, economy and timeliness.
Claims (3)
1. The bearing fault diagnosis method of the traction motor of the electric forklift in the complex noise environment is characterized by comprising the following steps:
step 1: acquiring a single-phase fault current data set I t={i1,t,i2,t,i3,t,...,in,t,...,iN,t of any one of phase A current, phase B current and phase C current of a traction motor of the electric forklift at the time T by using a phase current measurement link, wherein I n,t represents the nth single-phase current at the time T, and T is more than or equal to 0 and less than or equal to T s,Ts and is the total sampling time; n is more than 0 and less than or equal to N, wherein N is the total number of fault phase currents;
Let i n,t have a tag value of b n,t, where b e [1, D ], D is the number of failed species;
step 2: constructing a one-dimensional convolutional neural network, comprising: an input layer, K hidden layers and an output layer; wherein each hidden layer comprises: the device comprises a one-dimensional convolution unit, a compression unit, a linear rectification unit and a regularization unit;
Step 2.1: the input layer carries out convolution processing on the input i n,t to obtain a fault convolution characteristic i n,t *;
Step 2.2: when k=1, i n,t * is input into a kth hidden layer, a one-dimensional convolution unit processes i n,t * to obtain a kth information hidden state W n,t,k, a compression unit of the kth hidden layer processes the kth information hidden state W n,t,k to obtain a kth fault information feature X n,t,k, a linear rectification unit of the kth hidden layer processes X n,t,k to obtain a kth fault correction feature z n,t,k,zn,t,k, and the kth fault correction feature z n,t,k,zn,t,k is input into a regularization unit of the k hidden layer to process the kth fault information feature J n,t,k;
When k=2, 3, … and K, J n,t,k-1 inputs the K hidden layer for processing to obtain a K fault feature J n,t,k, and finally the K hidden layer outputs a K fault feature J n,t,K;
Step 2.3: the output layer processes J n,t,K to obtain a predicted tag value b n,t * of i n,t and outputs the predicted tag value b n,t *;
Step 4: based on b n,t and b n,t *, constructing a cross entropy loss function of a one-dimensional convolutional neural network, calculating the cross entropy loss function to update network parameters when training the convolutional neural network by using a gradient descent method, and stopping training until the cross entropy loss function converges, so as to obtain a trained electric fork truck traction motor bearing fault diagnosis model, and performing fault prediction on abnormal phase current when the electric fork truck traction motor works.
2. An electronic device comprising a memory and a processor, wherein the memory is configured to store a program that supports the processor to execute the bearing fault diagnosis method of claim 1, the processor being configured to execute the program stored in the memory.
3. A computer readable storage medium having a computer program stored thereon, characterized in that the computer program when run by a processor performs the steps of the bearing fault diagnosis method of claim 1.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410395053.1A CN118258606A (en) | 2024-04-02 | 2024-04-02 | Bearing fault diagnosis method for traction motor of electric forklift in complex noise environment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410395053.1A CN118258606A (en) | 2024-04-02 | 2024-04-02 | Bearing fault diagnosis method for traction motor of electric forklift in complex noise environment |
Publications (1)
Publication Number | Publication Date |
---|---|
CN118258606A true CN118258606A (en) | 2024-06-28 |
Family
ID=91605069
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202410395053.1A Pending CN118258606A (en) | 2024-04-02 | 2024-04-02 | Bearing fault diagnosis method for traction motor of electric forklift in complex noise environment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN118258606A (en) |
-
2024
- 2024-04-02 CN CN202410395053.1A patent/CN118258606A/en active Pending
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN116245256B (en) | Multi-factor-combined capacitor quality prediction method, system and storage medium | |
CN111579121B (en) | Method for diagnosing faults of temperature sensor in new energy automobile battery pack on line | |
US11372048B2 (en) | Diagnostic device and diagnostic method | |
CN111765075B (en) | Hydraulic forging press pump source fault prediction method and system | |
CN113792940A (en) | Hydro-turbo generator set runout data prediction method and device based on deep learning | |
CN115422759A (en) | Equipment fault diagnosis model construction and equipment fault diagnosis method and device | |
CN118258606A (en) | Bearing fault diagnosis method for traction motor of electric forklift in complex noise environment | |
CN116820821A (en) | Disk failure detection method, apparatus, electronic device and computer readable storage medium | |
CN117218602A (en) | Structural health monitoring data abnormality diagnosis method and system | |
CN111475966A (en) | Power electronic circuit fault diagnosis method based on digital twinning | |
CN116184229A (en) | Lithium ion battery internal short circuit fault diagnosis method and device based on voltage prediction | |
CN115962100A (en) | Wind turbine generator system state monitoring system | |
CN111044176A (en) | Method for monitoring temperature abnormity of generator | |
CN116342110B (en) | Intelligent fault diagnosis and fault tolerance measurement method for multiple temperature measurement loops of train | |
CN113835029B (en) | Motor fault diagnosis evaluation standard construction method, system and computer storage medium | |
CN117908522B (en) | Fault positioning method for initiating and controlling equipment based on simulation analysis | |
CN118194206B (en) | Intelligent fault diagnosis method and system for production equipment | |
CN117055537B (en) | Motorcycle electric control system fault diagnosis method based on Internet of things | |
CN116643170B (en) | Motor shafting vibration testing method and device and computer equipment | |
Liu et al. | A Continuous Evaluation Method and Model of Product Storage Lifetime | |
CN117194188A (en) | Method, system, electronic equipment and storage medium for server health diagnosis | |
CN118464459A (en) | Engine fault diagnosis method and device, cloud server and storage medium | |
CN116660757A (en) | Battery voltage estimation method, device and storage medium | |
CN117309396A (en) | Ball mill bearing fault prediction method and system based on self-encoder | |
CN117874663A (en) | Wind driven generator abnormality diagnosis device and diagnosis method thereof |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |