CN116663645A - Motor rotor system fault diagnosis method based on Sine-SSA-BP - Google Patents
Motor rotor system fault diagnosis method based on Sine-SSA-BP Download PDFInfo
- Publication number
- CN116663645A CN116663645A CN202310642459.0A CN202310642459A CN116663645A CN 116663645 A CN116663645 A CN 116663645A CN 202310642459 A CN202310642459 A CN 202310642459A CN 116663645 A CN116663645 A CN 116663645A
- Authority
- CN
- China
- Prior art keywords
- sparrow
- sine
- ssa
- population
- neural network
- 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 26
- 238000003745 diagnosis Methods 0.000 title claims abstract description 16
- 241000287127 Passeridae Species 0.000 claims abstract description 82
- 238000013528 artificial neural network Methods 0.000 claims abstract description 50
- 230000000739 chaotic effect Effects 0.000 claims abstract description 34
- 238000013507 mapping Methods 0.000 claims abstract description 23
- 238000012549 training Methods 0.000 claims abstract description 21
- 238000000605 extraction Methods 0.000 claims description 18
- 230000002431 foraging effect Effects 0.000 claims description 14
- 210000002569 neuron Anatomy 0.000 claims description 9
- 238000009826 distribution Methods 0.000 claims description 8
- 238000013459 approach Methods 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 6
- 238000010845 search algorithm Methods 0.000 claims description 5
- 230000004913 activation Effects 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000012544 monitoring process Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 238000012546 transfer Methods 0.000 claims description 3
- 238000004088 simulation Methods 0.000 abstract 1
- 238000005457 optimization Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 6
- 230000000694 effects Effects 0.000 description 3
- 238000012360 testing method Methods 0.000 description 2
- 230000002411 adverse Effects 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Classifications
-
- 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
- G06N3/08—Learning methods
- G06N3/086—Learning methods using evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- 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
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2123/00—Data types
- G06F2123/02—Data types in the time domain, e.g. time-series data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Evolutionary Biology (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Physiology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Control Of Electric Motors In General (AREA)
Abstract
The invention provides a motor rotor system fault diagnosis method based on a Sine-SSA-BP, which realizes the fault diagnosis of a motor rotor. 1) Determining the structure and related parameters of the BP neural network and extracting features; 2) Initializing sparrow population and position by using sine chaotic mapping; 3) The initial position of sparrow is used as weight and deviation of BP neural network; 4) Training a neural network by using the data of the training set, and obtaining an error value by forward propagation, wherein the error value is used as an adaptability value of the sparrow individual population; 5) Updating the sparrow position according to an SSA algorithm with improved sine chaotic mapping; 6) It is determined whether the algorithm satisfies a stop condition. If the stopping condition is met, outputting the optimal weight and deviation; if not, returning to the step 4 to continue executing the algorithm; 7) And taking the obtained optimal weight omega and deviation b as the final weight and deviation of the Sine-SSA-BP model, and carrying out motor rotor fault diagnosis by using the optimized Sine-SSA-BP model. Simulation results show that the optimized BP neural network has higher classification accuracy.
Description
Technical Field
The invention relates to the field of fault diagnosis, in particular to a motor rotor system fault diagnosis method based on a Sine-SSA-BP.
Background
The motor is used as an important power component of machinery, is widely applied to production and life, and plays an irreplaceable role. However, mechanical properties decrease with aging and life time of the coil wire, load change, working environment, and the like. This will inevitably lead to motor failure and even incur significant economic losses, life safety risks and other adverse effects. In fact, a significant portion of motor failure can switch the state of the rotor, an important component of the motor. Therefore, research on motor rotor system fault detection and diagnosis is very necessary and significant. By utilizing the global searching advantage of the genetic algorithm, the optimal weight and threshold of the BP neural network are optimized, so that the stability, generalization capability and convergence rate of the network can be greatly improved. However, if only the BP neural network is used, the global optimum is not easily obtained, the training frequency is increased, the learning efficiency is low, and the convergence speed is low. In order to solve the above problems, a sparrow search algorithm with improved sine chaotic mapping is provided to optimize the BP neural network, so as to improve the learning efficiency and convergence speed of the BP neural network.
Disclosure of Invention
The invention aims to: aiming at the problems in the background technology, the invention provides a motor rotor system fault diagnosis method based on the Sine-SSA-BP, and an improved sparrow optimization algorithm is used for optimizing the BP neural network, so that the learning efficiency and the convergence speed of the BP neural network are improved, and the motor rotor system fault diagnosis precision can be further improved.
The technical scheme is as follows: the invention discloses a motor rotor fault diagnosis method based on a Sine-SSA-BP, which comprises the following specific steps:
step 1: determining the structure and related parameters of the BP neural network, and performing feature extraction on an original motor vibration signal by adopting a time domain feature extraction method to obtain a motor rotor fault data set;
step 2: initializing a sparrow population and a sparrow position by using sine chaotic mapping, wherein the initial sparrow position is used as the weight and the deviation of the BP neural network;
step 3: training a BP neural network by using data of a training set, and obtaining an error value by forward propagation, wherein the error value is used as an adaptability value of a sparrow individual population;
step 4: updating the sparrow position according to the SSA algorithm with the improved sine chaotic mapping in the step 2;
step 5: determining whether an algorithm meets a stopping condition, and if so, outputting the optimal weight and deviation of the BP neural network; if not, returning to the step 3 to continue executing the algorithm;
step 6: and taking the obtained optimal weight omega and deviation v as the final weight and deviation of the Sine-SSA-BP model, and carrying out motor rotor fault diagnosis by using the optimized Sine-SSA-BP model.
Further, in the step 1, a time domain feature extraction method is adopted to perform feature extraction on an original motor vibration signal, and time domain information refers to a signal describing waveforms according to time, wherein the signal comprises dimension feature parameters and dimensionless feature parameters; adopting 6 dimension characteristic indexes and 7 dimensionless characteristic indexes to carry out characteristic extraction on original data of a motor rotor: the 6 dimension characteristic indexes comprise a mean value, std, RMS, a maximum value, a minimum value and a peak-to-peak value; the 7 dimensionless characteristic indexes comprise Crestfactor, shapefactor, implusefactor, energy and Skewness, kurtosis.
Further, initializing sparrow population and position by sine chaotic mapping in the step 2; the expression is as follows:
wherein x is k Indicating initial position of sparrow population, x k+1 Indicating the sparrow position after one iteration, a indicating the system parameters.
Further, the initial position of the sparrow in the step 2 is expressed by the following formula:
wherein omega ij Represented are connection weights between neurons,and θ is an activation function, a and b i For bias, i e (1, q) is the number of hidden layers, j e (1, m) is the number of neurons of the ith hidden layer.
Further, the error function E in the BP neural network in the step 3 is represented by the following formula:
wherein T represents actual output, T epsilon (1, mu) represents batch number of training samples, and Q is the selected optimal result after training by the BP neural network.
Further, in the step 4, the updating of the sparrow position by the SSA algorithm with improved sinusoidal chaotic mapping includes the following steps:
the first step: the sine chaotic map is used for initializing sparrow population, the proportion of discoverers and participants and the iteration times;
secondly, calculating the fitness value of each sparrow, and sequencing;
thirdly, updating the positions of the discoverers, the participants and the scouts by using a formula;
step four, calculating a fitness value and updating the sparrow position;
and fifthly, judging whether the preconditions are met, if the preconditions are met, exiting the algorithm and obtaining a group of optimal solutions, and if the preconditions are not met, continuing to repeat the second step to the fourth step.
Further, in the sparrow search algorithm, the sparrow population is divided into three groups, namely discoverers, participants and scouts; the position of each sparrow is a set of solution, and the sparrow with the highest fitness value can be obtained through calculation, and the position of the sparrow is the optimal solution needed by us;
the sparrows with higher fitness value are regarded as discoverers, play a role in searching food resources, provide the position and direction of food for the whole sparrow population, and update the formula of the position as follows:
wherein T represents the current iteration number, T max The number of iterations of the method is the maximum,represents the position of the ith sparrow at the t-th iteration, and beta 1 E (0, 1) is a random number, β 2 Random numbers subject to normal distribution, L representing a plurality of columns of a matrix, R 2 E (0, 1) is an alarm value, ST E (0.5, 1) represents a security threshold; when R is 2 When ST is not less than, the natural enemy possibly appears near the population, and the whole population needs to fly to a safe area to continue foraging; r is R 2 The ST is less than that the current foraging environment is very safe, discoverers can conduct extensive searching, and population life is guided to a higher fitness value;
during the foraging process, the participants always follow the discoverers, and once the discoverers find food, the participants compete with the discoverers for food; if the participant fails, the participant will continue to fly elsewhere to continue to feed, and the location of the participant's update formula may be expressed as follows:
wherein X is worst Representing the current global worst of the ith iterationA location; x is X P Representing the best position of the finder in the ith iteration; a represents a row multi-column matrix, and all elements are 1 or-1; a is that + ＝A T (AA T ) -1 The method comprises the steps of carrying out a first treatment on the surface of the When (when)This means that the participants in this section do not get food in competition with the discoverer, they need to fly to other areas to continue to find food, or else the participants get food in competition with the discoverer;
when the whole population forges, the scout is responsible for monitoring the whole foraging area, and the scout can abandon foraging and transfer to a safe area or randomly approach other sparrows so as to reduce the risk of predation when enemies approach, and the position update formula of the scout can be expressed as:
wherein X is best Representing the global optimal position of the current population, beta 3 The step control parameter for controlling the direction and distance of sparrow movement is a random number following standard normal distribution, beta 4 ∈[-1，1]Is a random number, ρ is a constant, the denominator of zero is avoided, f i Indicating the fitness value of the ith sparrow, f g And f ω And the fitness values representing the global optimal position and the global worst position respectively.
The beneficial effects are that:
1. the chaotic mapping initializing sparrow population can effectively improve the uniformity of initial population distribution and avoid sinking into a local optimal solution. The sinusoidal chaotic map is a model with an infinite number of map folds as one of the chaotic maps. Compared with the ent and logic chaotic mapping, the sine chaotic mapping has better chaotic characteristics, the sine chaotic mapping is selected to initialize sparrow populations in SSA, and the random property of the sine chaotic mapping is utilized to enhance the global searching capability of a particle swarm algorithm, so that the situation of sinking into local optimum is avoided.
2. The invention adopts a time domain feature extraction method to extract the features of the signal, the method mainly obtains information from the waveform, can effectively improve the signal-to-noise ratio, obtains the similarity and the relevance of the waveform of the signal at different moments, obtains the feature parameters reflecting the running state of the mechanical equipment, and provides effective information for dynamic analysis and fault diagnosis of a mechanical system.
3. According to the invention, the BP neural network is optimized by adopting a sine chaotic mapping optimized sparrow population optimization method, the output of the optimized BP neural network is basically consistent with the actual output, and the output of the non-optimized BP and the actual output still have a plurality of larger deviations, so that the error between the output value and the true value of the optimized BP neural network is closer to 0. Compared with an unoptimized BP neural network, the BP neural network subjected to the Sine-SSA optimization has more accurate classification precision.
Drawings
FIG. 1 is a general flow chart of the present invention;
FIG. 2 is a graph showing the evolutionary convergence of the Sine-SSA of the present invention;
FIG. 3 is a graph showing the comparison of predicted and actual values of BP neural network before and after the optimization of the Sine-SSA in the present invention;
FIG. 4 is a graph showing the error between the predicted value and the true value of the BP neural network before and after the SIne-SSA optimization of the present invention.
Detailed description of the preferred embodiments
The invention is further described below with reference to the accompanying drawings. The following examples are only for more clearly illustrating the technical aspects of the present invention, and are not intended to limit the scope of the present invention.
The invention discloses a motor rotor system fault diagnosis method based on a Sine-SSA-BP, which is shown in fig. 1 and specifically comprises the following steps:
step 1: the method comprises the steps of determining the structure and related parameters of the BP neural network, and extracting the characteristics of the original motor vibration signals by adopting a time domain characteristic extraction method. Since the information obtained from the original vibration signal is not sufficiently obvious, the signal is subjected to feature extraction by a time domain feature extraction method.
The invention adopts 6 dimension characteristic indexes and 7 dimensionless characteristic indexes to carry out characteristic extraction on the original data of machine vibration signals, and the characteristic extraction is specifically shown in a table 1.
TABLE 1 dimension characteristic index
The partial data after time domain feature extraction are shown in table 2:
table 2 time domain feature extraction data
In the table, the first 13 columns of data are corresponding characteristic values calculated for the original vibration signals according to the formula of 13 characteristic indexes in the upper graph, the 14 th column of data are labels of motor rotor states, if the labels are 1, the motor rotor is not faulty, and if the labels are 2, the motor rotor is faulty.
Step 2: the sparrow population and the position are initialized by using the sine chaotic map, and the sparrow population is initialized by using the chaotic map, so that the uniformity of the initial population distribution can be effectively improved, and the sparrow population is prevented from falling into a local optimal solution. The sinusoidal chaotic map is a model with an infinite number of map folds as one of the chaotic maps. Compared with the ent and logics chaotic mapping, the sine chaotic mapping has better chaotic characteristics, so the sine chaotic mapping is selected to initialize the sparrow population in the SSA, and the mathematical expression of the sine chaotic mapping is as follows in the formula (1):
wherein x is k Indicating initial position of sparrow population, x k+1 Indicating the sparrow position after one iteration, a indicating the system parameters.
Step 3: the initial position of sparrow is used as weight and deviation of BP neural network, and the specific operation steps are as follows:
1. the learning algorithm of the BP neural network comprises forward propagation and backward propagation. During forward propagation, the input of the ith neuron of the hidden layer may be expressed as equation (2):
the output of the neuron can be expressed as:
neuron input through the output layer after the hidden layer can be expressed as follows:
the final output result of the BP neural network can be expressed as:
wherein omega ij Represented are connection weights between neurons,and θ is an activation function, a and b i For bias, I e (1, q) is the number of hidden layers, j e (1, m) is the number of neurons of the I hidden layer.
Step 4: the neural network is trained by the data of the training set, an error value is obtained through forward propagation, the error value is used as the fitness value of the sparrow individual population, and the back propagation of the BP neural network is used for adjusting the weight and the deviation according to the error between the actual output and the calculated output of the neural network. Herein, the error function E in the network can be represented by the following equation (6):
wherein T represents actual output, T epsilon (1, mu) represents batch number of training samples, and Q is the selected optimal result after training by the BP neural network.
Step 5: and updating the sparrow position according to the SSA algorithm with the improved sine chaotic mapping.
The first step: sinusoidal chaotic map is used to initialize sparrow population, the ratio of discoverers and participants, and the number of iterations.
And a second step of: and calculating the fitness value of each sparrow, and sequencing.
And a third step of: the location of the discoverers, participants, and scouts is updated using the formulas.
Fourth step: and calculating the fitness value and updating the sparrow position.
Fifth step: judging whether the precondition is met, if the precondition is met, exiting the algorithm and obtaining a group of optimal solutions, and if the precondition is not met, continuing to repeat the steps 2 to 4.
In the sparrow search algorithm, the sparrow population is divided into three groups, discoverers, participants and scouts. The location of each sparrow is a solution. Finally, the sparrow with the highest fitness value can be obtained through calculation, and the position of the sparrow is the optimal solution needed by people.
The sparrows with higher fitness value are regarded as discoverers, play a role in searching food resources, provide the position and direction of food for the whole sparrow population, and update the formula (7) of the position as follows:
wherein T represents the current iteration number, T max The number of iterations of the method is the maximum,represents the position of the ith sparrow at the t-th iteration, and beta 1 E (0, 1) is a random number, β 2 Random numbers subject to normal distribution, L representing a plurality of columns of a matrix, R 2 E (0, 1) is an alarm value, ST E (0.5, 1) represents a security threshold; when R is 2 When ST is not less than, the natural enemy possibly appears near the population, and the whole population needs to fly to a safe area to continue foraging; r is R 2 < ST indicates that the current foraging environment is very safe, discoverers can conduct extensive searches, and population life is guided to a higher fitness value.
During the foraging process, participants always follow discoverers. Once the discoverer finds food, the participant competes with the discoverer for food. If the participant fails, the participant will continue to fly elsewhere to continue to feed. The location of the participant update formula is as in formula (8):
wherein X is worst Representing the current global worst position of the ith iteration; x is X P Representing the best position of the finder in the ith iteration; a represents a row multi-column matrix, and all elements are 1 or-1; a is that + ＝A T (AA T ) -1 The method comprises the steps of carrying out a first treatment on the surface of the When (when)This means that the participants in this section do not get food in competition with the discoverer, they need to fly to other areas to continue to find food, or else the participants get food in competition with the discoverer;
when the whole population forges, the scout is responsible for monitoring the whole foraging area, and the scout can abandon foraging and transfer to a safe area or randomly approach other sparrows so as to reduce the risk of predation when enemies approach, and the position update formula of the scout can be expressed as:
wherein X is best Representing the global optimal position of the current population, beta 3 The step control parameter for controlling the direction and distance of sparrow movement is a random number following standard normal distribution, beta 4 ∈[-1，1]Is a random number, ρ is a constant, the denominator of zero is avoided, f i Indicating the fitness value of the ith sparrow, f g And f w And the fitness values representing the global optimal position and the global worst position respectively.
Step 6: it is determined whether the algorithm satisfies a stop condition. If the stopping condition is met, outputting the optimal weight and deviation; if not, returning to the step 4, and continuing to execute the algorithm by using the formula (6);
when SSA is used to optimize the initial weight ω and the bias b of BP, the best fitness value of the sparrow population in each iteration is shown in fig. 2, and in the sparrow search algorithm, the minimum classification error rate of the training set is taken as the fitness function, so as to find a set of weight ω and bias b to minimize the fitness function value. The curve shown in the above figure is a gradually decreasing curve.
Step 7: experiments prove that firstly, according to the dimension of input data and output data, the number of input nodes of the BP neural network is 13, and the number of output nodes is 1. The following empirical formula is then used to determine the number of nodes in the hidden layer of the BP neural network:
hiddennum＝sqrt(m+n)+a
where m and n are the number of nodes of the input layer and the output layer, respectively, and a is typically an integer between 1 and 10. According to the above formula, the upper bound of the hidden layer node number is determined to be 13, and then the hidden layer node number with the minimum error is obtained by using a training set. The program operation results are shown in table 3:
table 3 mean square error of training set
| Sequence number | Mean square error of training set | Sequence number | Mean square error of training set |
| 4 | 1.2023e-06 | 9 | 3.0518e-07 |
| 5 | 3.503e-06 | 10 | 3.7e-07 |
| 6 | 1.5032e-06 | 11 | 1.2576e-06 |
| 7 | 1.2528e-06 | 12 | 1.4292e-07 |
| 8 | 3.4105e-08 | 13 | 1.457e-06 |
As can be seen from the above table, the mean square error of the training set is minimal when the number of hidden layer nodes is 8. Therefore, the hidden layer node number of the BP neural network is set to 8 herein.
After the structure of the BP neural network is determined, the training set data are respectively brought into the non-optimized BP neural network and the optimized BP neural network. After the corresponding weights ω and bias b are obtained, the test set is brought into test. The classification result is shown in fig. 3. In the figure, asterisks represent true output, triangles represent output of the BP neural network, and black circles represent output of the SSA optimized BP neural network with improved sinusoidal chaotic mapping. From the graph, the output of the optimized BP is basically consistent with the actual output, while the output of the non-optimized BP has a few larger deviations from the actual output, so that the optimized BP has better effect.
In order to more intuitively observe the superiority of the optimized BP neural network, the optimized BP output value and the non-optimized BP output value are respectively compared with the real output to obtain an error comparison graph of the output value and the real value of the Sine-SSA model before and after optimization, as shown in fig. 4. As can be seen from the figure, compared with the non-optimized BP neural network, the error between the output value and the true value of the optimized BP neural network is closer to 0, so that the classification effect of the BP after the Sine-SSA optimization is far better than that of the non-optimized BP. In summary, the more accurate classification accuracy of the Sine-SSA optimized BP neural network is achieved compared to the non-optimized BP neural network.
The foregoing embodiments are merely illustrative of the technical concept and features of the present invention, and are intended to enable those skilled in the art to understand the present invention and to implement the same, not to limit the scope of the present invention. All equivalent changes or modifications made according to the spirit of the present invention should be included in the scope of the present invention.
Claims (7)
1. A motor rotor fault diagnosis method based on the Sine-SSA-BP is characterized by comprising the following specific steps:
step 1: determining the structure and related parameters of the BP neural network, and performing feature extraction on an original motor vibration signal by adopting a time domain feature extraction method to obtain a motor rotor fault data set;
step 2: initializing a sparrow population and a sparrow position by using sine chaotic mapping, wherein the initial sparrow position is used as the weight and the deviation of the BP neural network;
step 3: training a BP neural network by using data of a training set, and obtaining an error value by forward propagation, wherein the error value is used as an adaptability value of a sparrow individual population;
step 4: updating the sparrow position according to the SSA algorithm with the improved sine chaotic mapping in the step 2;
step 5: determining whether an algorithm meets a stopping condition, and if so, outputting the optimal weight and deviation of the BP neural network; if not, returning to the step 3 to continue executing the algorithm;
step 6: and taking the obtained optimal weight omega and deviation b as the final weight and deviation of the Sine-SSA-BP model, and carrying out motor rotor fault diagnosis by using the optimized Sine-SSA-BP model.
2. The method for diagnosing motor rotor faults based on the Sine-SSA-BP according to claim 1, wherein in the step 1, a time domain feature extraction method is adopted to perform feature extraction on an original motor vibration signal, and time domain information refers to a signal describing waveforms according to time, and the signal comprises dimensional feature parameters and dimensionless feature parameters; adopting 6 dimension characteristic indexes and 7 dimensionless characteristic indexes to carry out characteristic extraction on original data of a motor rotor: the 6 dimension characteristic indexes comprise a mean value, std, RMS, a maximum value, a minimum value and a peak-to-peak value; the 7 dimensionless characteristic indexes comprise Crestfactor, shapefactor, implusefactor, energy and Skewness, kurtosis.
3. The method for diagnosing motor rotor faults based on the Sine-SSA-BP according to claim 1, wherein the sinusoidal chaotic mapping in the step 2 initializes the sparrow population and the position; the expression is as follows:
wherein x is k Indicating initial position of sparrow population, x k+1 Indicating the sparrow position after one iteration, a indicating the system parameters.
4. The method for diagnosing motor rotor failure based on the fine-SSA-BP according to claim 1, wherein the initial position of the sparrow in the step 2 is expressed by the following formula:
wherein omega ij Represented are connection weights between neurons,and θ is an activation function, a and b i For bias, i e (1, q) is the number of hidden layers, j e (1, m) is the number of neurons of the ith hidden layer.
5. The method for diagnosing motor rotor failure based on the Sine-SSA-BP according to claim 1, wherein the error function E in the BP neural network in the step 3 is represented by the following formula:
wherein T represents actual output, T epsilon (1, mu) represents batch number of training samples, and Q is the selected optimal result after training by the BP neural network.
6. The method for diagnosing motor rotor failure based on Sine-SSA-BP according to claim 1, wherein in step 4, updating sparrow positions by the SSA algorithm with improved sinusoidal chaotic mapping comprises the steps of:
the first step: the sine chaotic map is used for initializing sparrow population, the proportion of discoverers and participants and the iteration times;
secondly, calculating the fitness value of each sparrow, and sequencing;
thirdly, updating the positions of the discoverers, the participants and the scouts by using a formula;
step four, calculating a fitness value and updating the sparrow position;
and fifthly, judging whether the preconditions are met, if the preconditions are met, exiting the algorithm and obtaining a group of optimal solutions, and if the preconditions are not met, continuing to repeat the second step to the fourth step.
7. The method for diagnosing motor rotor faults based on fine-SSA-BP of claim 6, wherein in the sparrow search algorithm, the sparrow population is divided into three groups, discoverers, participants and scouts; the position of each sparrow is a set of solution, and the sparrow with the highest fitness value can be obtained through calculation, and the position of the sparrow is the optimal solution needed by us;
the sparrows with higher fitness value are regarded as discoverers, play a role in searching food resources, provide the position and direction of food for the whole sparrow population, and update the formula of the position as follows:
wherein T represents the current iteration number, T max The number of iterations of the method is the maximum,represents the position of the ith sparrow at the t-th iteration, and beta 1 E (0, 1) is a random number, β 2 Random numbers subject to normal distribution, L representing a plurality of columns of a matrix, R 2 E (0, 1) is an alarm value, ST E (0.5, 1) represents a security threshold; when R is 2 Meaning when ST is not less thanNatural enemies may appear near the population, and the whole population needs to fly to a safe area to continue to find food; r is R 2 The ST is less than that the current foraging environment is very safe, discoverers can conduct extensive searching, and population life is guided to a higher fitness value;
during the foraging process, the participants always follow the discoverers, and once the discoverers find food, the participants compete with the discoverers for food; if the participant fails, the participant will continue to fly elsewhere to continue to feed, and the location of the participant's update formula may be expressed as follows:
wherein X is worst Representing the current global worst position of the ith iteration; x is X P Representing the best position of the finder in the ith iteration; a represents a row multi-column matrix, and all elements are 1 or-1; a is that + ＝A T (AA T ) -1 The method comprises the steps of carrying out a first treatment on the surface of the When (when)This means that the participants in this section do not get food in competition with the discoverer, they need to fly to other areas to continue to find food, or else the participants get food in competition with the discoverer;
when the whole population forges, the scout is responsible for monitoring the whole foraging area, and the scout can abandon foraging and transfer to a safe area or randomly approach other sparrows so as to reduce the risk of predation when enemies approach, and the position update formula of the scout can be expressed as:
wherein x is best Representing the global optimal position of the current population, beta 3 The step control parameter for controlling the direction and distance of sparrow movement is a random number following standard normal distribution，β 4 ∈[-1，1]Is a random number, ρ is a constant, the denominator of zero is avoided, f i Indicating the fitness value of the ith sparrow, f g And f w And the fitness values representing the global optimal position and the global worst position respectively.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310642459.0A CN116663645A (en) | 2023-06-01 | 2023-06-01 | Motor rotor system fault diagnosis method based on Sine-SSA-BP |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310642459.0A CN116663645A (en) | 2023-06-01 | 2023-06-01 | Motor rotor system fault diagnosis method based on Sine-SSA-BP |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116663645A true CN116663645A (en) | 2023-08-29 |
Family
ID=87723790
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310642459.0A Pending CN116663645A (en) | 2023-06-01 | 2023-06-01 | Motor rotor system fault diagnosis method based on Sine-SSA-BP |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116663645A (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108376286A (en) * | 2018-02-26 | 2018-08-07 | 上海理工大学 | A kind of equipment fault diagnosis apparatus and system |
| US20190242936A1 (en) * | 2018-02-05 | 2019-08-08 | Wuhan University | Fault diagnosis method for series hybrid electric vehicle ac/dc converter |
| US20190243735A1 (en) * | 2018-02-05 | 2019-08-08 | Wuhan University | Deep belief network feature extraction-based analogue circuit fault diagnosis method |
| CN110232435A (en) * | 2019-04-30 | 2019-09-13 | 沈阳化工大学 | A kind of adaptive depth confidence network Fault Diagnosis of Roller Bearings |
| CN113157732A (en) * | 2021-02-23 | 2021-07-23 | 北京科技大学 | Underground scraper fault diagnosis method based on PSO-BP neural network |
| CN114565259A (en) * | 2022-02-24 | 2022-05-31 | 武汉大学 | Power distribution network line loss calculation method for optimizing BP neural network by improving sparrow search |
| CN114925612A (en) * | 2022-05-27 | 2022-08-19 | 江苏大学 | Transformer fault diagnosis method for optimizing hybrid kernel extreme learning machine based on sparrow search algorithm |
| CN115249217A (en) * | 2022-01-17 | 2022-10-28 | 浙江理工大学 | Spectral reflectivity reconstruction method based on improved BP neural network |
| CN115982575A (en) * | 2023-02-27 | 2023-04-18 | 天津智能轨道交通研究院有限公司 | Rolling bearing fault diagnosis method and device, electronic equipment and readable storage equipment |
-
2023
- 2023-06-01 CN CN202310642459.0A patent/CN116663645A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190242936A1 (en) * | 2018-02-05 | 2019-08-08 | Wuhan University | Fault diagnosis method for series hybrid electric vehicle ac/dc converter |
| US20190243735A1 (en) * | 2018-02-05 | 2019-08-08 | Wuhan University | Deep belief network feature extraction-based analogue circuit fault diagnosis method |
| CN108376286A (en) * | 2018-02-26 | 2018-08-07 | 上海理工大学 | A kind of equipment fault diagnosis apparatus and system |
| CN110232435A (en) * | 2019-04-30 | 2019-09-13 | 沈阳化工大学 | A kind of adaptive depth confidence network Fault Diagnosis of Roller Bearings |
| CN113157732A (en) * | 2021-02-23 | 2021-07-23 | 北京科技大学 | Underground scraper fault diagnosis method based on PSO-BP neural network |
| CN115249217A (en) * | 2022-01-17 | 2022-10-28 | 浙江理工大学 | Spectral reflectivity reconstruction method based on improved BP neural network |
| CN114565259A (en) * | 2022-02-24 | 2022-05-31 | 武汉大学 | Power distribution network line loss calculation method for optimizing BP neural network by improving sparrow search |
| CN114925612A (en) * | 2022-05-27 | 2022-08-19 | 江苏大学 | Transformer fault diagnosis method for optimizing hybrid kernel extreme learning machine based on sparrow search algorithm |
| CN115982575A (en) * | 2023-02-27 | 2023-04-18 | 天津智能轨道交通研究院有限公司 | Rolling bearing fault diagnosis method and device, electronic equipment and readable storage equipment |
Non-Patent Citations (3)
| Title |
|---|
| LIANG ZHANG ET AL.: "Research on neural network wind speed prediction model based on improved sparrow algorithm optimization", 《2022 INTERNATIONAL CONFERENCE ON ENERGY STORAGE TECHNOLOGY AND POWER SYSTEMS (ESPS 2022)》, vol. 8, 18 October 2022 (2022-10-18), pages 739 - 747 * |
| ZOU DINGJIANG ET AL.: "Power Load Forecasting Based on Sine-SSA-BP Neural Network", 《CONFERENCE PROCEEDINGS OF 2021 INTERNATIONAL JOINT CONFERENCE ON ENERGY, ELECTRICAL AND POWER ENGINEERING》, vol. 899, 3 September 2022 (2022-09-03), pages 565 - 572 * |
| 王晋: "基于小波包、BP神经网络和SVM的滚动轴承故障诊断研究", 《中国优秀硕士学位论文全文数据库 (工程科技Ⅱ辑)》, 15 February 2023 (2023-02-15), pages 029 - 815 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114969995B (en) | Rolling bearing early fault intelligent diagnosis method based on improved sparrow search and acoustic emission | |
| CN108062572A (en) | A kind of Fault Diagnosis Method of Hydro-generating Unit and system based on DdAE deep learning models | |
| CN110082136B (en) | Rotary machine fault diagnosis method based on cloud genetic algorithm optimization support vector machine | |
| CN113190931A (en) | Sub-health state identification method for improving optimized DBN-ELM of wolf | |
| CN109190270A (en) | A kind of double balancing disk balance Control Scheme methods based on APSO-BP | |
| CN111624522A (en) | Ant colony optimization-based RBF neural network control transformer fault diagnosis method | |
| CN108594793A (en) | A kind of improved RBF flight control systems fault diagnosis network training method | |
| CN112149883A (en) | Photovoltaic power prediction method based on FWA-BP neural network | |
| CN115859077A (en) | Multi-feature fusion motor small sample fault diagnosis method under variable working conditions | |
| CN112132102A (en) | Intelligent fault diagnosis method combining deep neural network with artificial bee colony optimization | |
| CN113221447A (en) | Soil humidity prediction method for optimizing BP neural network based on improved genetic algorithm | |
| CN113687654A (en) | Neural network training method and path planning method based on evolutionary algorithm | |
| CN116881841A (en) | Hybrid model fault diagnosis method based on F1-score multistage decision analysis | |
| CN117574057B (en) | Intelligent fault diagnosis method for vertical water pump unit | |
| CN113095477B (en) | Wind power prediction method based on DE-BP neural network | |
| Liang et al. | Recent advances in particle swarm optimization via population structuring and individual behavior control | |
| CN114037050A (en) | Robot degradation environment obstacle avoidance method based on internal plasticity of pulse neural network | |
| Addeh et al. | Application of the PSO-RBFNN model for recognition of control chart patterns | |
| CN116663645A (en) | Motor rotor system fault diagnosis method based on Sine-SSA-BP | |
| CN110363232B (en) | BP neural network-based millimeter wave detector interference effect evaluation method | |
| CN116680969A (en) | Filler evaluation parameter prediction method and device for PSO-BP algorithm | |
| Zhang et al. | Improved Probabilistic Neural Network Based Fault Diagnosis of Control Valve | |
| Wu et al. | An Improved FOA to Optimize GRNN Method for Wind Turbine Fault Diagnosis. | |
| Di et al. | Application of Genetic Algorithm and Convolutional Neural Network in Mobile Robot Path Planning | |
| Zhang et al. | Star Identification Utilizing Neural Networks |
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 |