CN109977869B - Shore bridge running mechanism state evaluation method based on deep learning - Google Patents
Shore bridge running mechanism state evaluation method based on deep learning Download PDFInfo
- Publication number
- CN109977869B CN109977869B CN201910235249.3A CN201910235249A CN109977869B CN 109977869 B CN109977869 B CN 109977869B CN 201910235249 A CN201910235249 A CN 201910235249A CN 109977869 B CN109977869 B CN 109977869B
- Authority
- CN
- China
- Prior art keywords
- shore bridge
- state
- deep learning
- layer
- running mechanism
- 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.)
- Active
Links
Images
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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a shore bridge running mechanism state evaluation method based on deep learning, which comprises the following steps: firstly, collecting vibration signals of a key operation mechanism in a shore bridge machine room, calculating relevant characteristic indexes of the collected vibration signals, then adaptively extracting signal state characteristic indexes based on building a trained 1D convolutional neural network model, outputting results in a grading mode, and evaluating the state of the shore bridge operation mechanism according to the size and stability of the scores. The method solves the problem that expert experience is needed for manually extracting the characteristic indexes, and the accuracy of state identification of the shore bridge running mechanism is effectively improved by intelligently processing the characteristic indexes. Meanwhile, evaluation is carried out in a scoring mode, and the health state of the shore bridge running mechanism is visually reflected.
Description
Technical Field
The invention belongs to the technical field of state monitoring of mechanical equipment, and particularly relates to an intelligent monitoring and state evaluation method for a bank bridge running mechanism, which is used for monitoring and evaluating a key running mechanism of a bank bridge in an automatic wharf.
Technical Field
With the increasing frequency of trade, the requirements for the loading and unloading capacity and efficiency of container terminals are continuously increased, large container loading and unloading equipment (shore bridges) in ports play an important role, and shore container cranes (shore bridges) are complex in structure, various in loading action form, complex and severe in operation conditions, in particular key operation mechanisms of the shore bridges, such as: the lifting mechanism, the pitching mechanism, the trolley running mechanism, the reduction gearbox and the like are frequently operated under the working conditions of continuous large load, large impact and frequent start and stop, various faults are easy to occur in the running process, once the faults occur, huge economic loss can be caused, even the normal running of the whole shore bridge system is influenced, and therefore intelligent health state assessment is necessary to be carried out on the key running mechanism of the shore bridge.
With the progress of sensor technology and communication technology and the rise of cloud computing technology, the number, scale and variety of detection signals of machine equipment are continuously increased, so that a foundation is provided for intelligent state evaluation of the machine equipment, but no intelligent state evaluation method specially used for a shore bridge exists at present, and for a traditional shore bridge state evaluation method, some problems exist:
(1) Most evaluation methods require manual selection of key feature indicators, such as wavelet packet transformation methods, and although these methods have been proven to successfully extract the required features and perform state evaluation, the upper limit performance of these methods depends on the quality of the feature design, it is difficult to pre-design manual features, it requires rich signal processing knowledge, and the adaptability of the manually designed features to newly measured data is poor.
(2) Most of evaluation modes are a mode of setting threshold value alarm (alarm mode when acquired data exceeds a certain numerical value) to achieve the purpose of fault alarm, if the threshold value is improperly set, the alarm is easily missed and false, but the threshold value is difficult to accurately set, meanwhile, the fixed threshold value has poor adaptability to different states, and the instability of a system is greatly improved due to simple threshold value alarm.
In order to solve the problems, the invention provides a shore bridge running mechanism state evaluation method based on deep learning. Compared with the traditional method based on classical machine learning, the intelligent state monitoring method based on deep learning abandons artificial experience intervention in signal feature extraction, utilizes the deep learning method to process input data layer by layer, gradually converts initial low-level features which are not related to the health state of the shore bridge mechanism into high-level features which are closely related to the health state of equipment, and realizes the self-adaptive extraction of the state features of the shore bridge mechanism; meanwhile, the feature extraction and state recognition processes are combined into a whole, the nonlinear mapping relation between the high-level features and the health state of the shore bridge mechanism is directly established, an intelligent state evaluation model is obtained, and the result is output in a grading mode.
Disclosure of Invention
In order to effectively solve the problems and promote the intelligent evaluation of the health state of the key operating mechanism of the shore bridge, the invention provides a method for evaluating the state of the key operating mechanism of the shore bridge based on deep learning, which specifically comprises the following steps:
s1, signal acquisition: an acceleration sensor is used for acquiring vibration signals of key operating mechanisms (such as a hoisting mechanism, a pitching mechanism, a trolley operating mechanism and the like) of the shore bridge.
S2, calculating a characteristic index: calculating the common characteristic indexes of the vibration signals of the shore bridge running mechanism, wherein the common characteristic indexes comprise: effective value, mean value, standard deviation, vibration intensity, kurtosis, skewness and frequency spectrum.
S3, training a model: and (3) building a 1D Convolutional Neural Network (CNN) based on a deep learning algorithm, selecting vibration signal data of the labeled quayside crane operating mechanism as a training set, learning by using the vibration characteristic indexes calculated in the step (S2), and training a convolutional Neural Network model.
Step S4, state evaluation: inputting the collected real-time vibration signal of the operating mechanism in the shore bridge machine room, extracting the characteristic index of the vibration signal according to the step S2, outputting the score of the current state of the shore bridge operating mechanism by using the model built in the step S3, and evaluating the health state of the shore bridge operating mechanism according to the score.
Further, the step S2 of extracting the characteristic index of the acquired vibration signal data by using a vibration multi-statistics time domain frequency domain characteristic extraction algorithm includes the following steps:
s21, calculating an effective value X rms A characteristic index reflecting the average energy level of the signal dynamic and static,
s22, calculating a mean value X m A characteristic index reflecting the central position of the signal and the average level of change,
s23, calculating targetTolerance X std A characteristic index reflecting the degree of fluctuation of the signal at the center position,
s24, calculating the characteristic index of the vibration intensity T, reflecting the energy,
s25, calculating kurtosis X kur Characteristic indexes, pulse vibration degree and impact energy,
s26, calculating a skewness X ske The characteristic index can reflect the sensitivity to the vibration signal,
wherein, X i Is a data value; x m Is the average of the data; n is a data point;
and S27, calculating the frequency spectrum of the vibration signal, and obtaining the frequency spectrum through Fast Fourier Transform (FFT) calculation, wherein the frequency composition structure and the amplitude of the vibration signal can be reflected.
Further, step S3 is to train the model, and vibration signal data of the shore bridge running mechanism with the label is selected as a training set; calculating a data characteristic index by using the step S2; based on a deep learning algorithm, the data characteristic indexes obtained through calculation are used as convolutional neural network input, the output value of each neuron is calculated in the forward direction, the error with the actual value is calculated, then the error is propagated in the backward direction, the gradient of the connection weight of each neuron is calculated, the weight of each characteristic index is updated according to a gradient descent rule, when the set working state identification accuracy is achieved, training is finished, and a model with the capability of grading and estimating the state of the quayside container crane operating mechanism is obtained.
Preferably, a 1D convolutional neural network is used, the network comprising convolutional layers, pooling layers, fully-connected layers; inputting the vibration signal characteristic indexes of the shore bridge operating mechanism in a matrix form, and performing convolution operation on a convolution layer to extract characteristics from the convolution layer; performing pooling operation on a pooling layer to reduce data dimensionality and simultaneously performing secondary feature extraction to avoid overfitting; and (3) converting the bottom layer characteristics into more abstract high-layer characteristics layer by layer through multilayer convolution and pooling treatment, outputting the result to a full connection layer, integrating the characteristics extracted by the convolution layer and the pooling layer, and outputting the high-level characteristics of the data.
Preferably, the neuron output of the last fully-connected layer is a quayside state score. Using Sigmoid activation function, the formula is:
wherein, W i Is a convolution kernel weight matrix;
is a characteristic index of the l-th layer; t is a Sigmoid function input value, a function output value S (t) is between 0 and 100, and the result is converted into a score output of between 0 and 100 through a Sigmoid activation function of linear regression.
Further, convolutional neural network training belongs to supervised learning, the error is calculated by adopting an MSE loss function and an L2 regularization method, and the formula is as follows:
wherein h is W,b (x (i) ) Outputting the score of the shore bridge running mechanism by the convolutional neural network; y is (i) Scoring the actual labels of the quayside container crane operating mechanism; i number of samples (i =1,2, \8230;, m);
is a convolution kernel weight matrix; b is a mixture of (i) Is the grid deviation; λ is a regularization parameter.
Further, the error is propagated from output layer to input layer by layer to carry out back propagation, the gradient of the error to the layer parameter is calculated, the weight of each characteristic index is updated according to a gradient descent rule, the training set is used for continuous training, the weight of the loss minimization function is found, the error minimization can be realized, when the set working state identification accuracy is reached, the training is finished, and the training model with better performance is obtained.
Further, in the step S4, the state evaluation is performed, after the data characteristic index is calculated according to the step S2, the acquired real-time vibration signal of the operating mechanism in the quay crane machine room is input into the model trained in the step S3, the score of the current state of the quay crane operating mechanism is output, and the health state and the stability degree of the quay crane operating mechanism are obtained through the analysis of the score; the lower the score of the vibration signal characteristic index is, the smaller the probability that the signal is in a normal state is, and meanwhile, the score change is small, so that the machine is in a stable state, and the shore bridge state can be effectively evaluated by the model.
The invention has the following beneficial effects:
(1) According to the shore bridge running mechanism state evaluation method based on deep learning, relevant data of a key shore bridge running mechanism are collected, a training convolutional neural network model is built by applying a vibration multi-statistics time-domain frequency-domain feature extraction algorithm, real-time shore bridge data are intelligently evaluated based on the trained model, the state of the shore bridge running mechanism is visually displayed in a grading mode, more accurate evaluation of the state of the shore bridge running mechanism is obtained, and the method is beneficial for port machinery to develop towards the direction of automation, no-man and intelligence.
(2) According to the state evaluation method of the shore bridge running mechanism based on deep learning, provided by the invention, the autonomous selection of time domain characteristic indexes is realized through the mapping and conversion process of multiple hidden layers in a convolutional neural network, the adaptive characteristic extraction of frequency domain data is realized, and further, the characteristic indexes are fused, so that the shore bridge state is more accurately described by deep effective characteristic indexes, and the problem that the characteristic indexes are difficult to select and design manually is effectively solved.
(3) The invention relates to a shore bridge running mechanism state evaluation method based on deep learning, which adopts a new evaluation formula intelligent evaluation standard, uses a Sigmoid function to quantify the health state evaluation standard of a shore bridge running mechanism, outputs a result by a score of 0 to 100, can know the health state of the shore bridge running mechanism according to the output score, and effectively solves the problem that a threshold value cannot be accurately set.
Drawings
In order to more clearly illustrate the embodiments or technical solutions of the present invention, the drawings used in the embodiments or technical solutions in the prior art are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a data flow schematic diagram of a state evaluation method of a shore bridge running mechanism based on deep learning.
FIG. 2 is a schematic diagram of a convolutional neural network structure based on a deep learning shore bridge operating mechanism state evaluation method.
FIG. 3 is a flow chart of an implementation of a deep learning based quayside container crane operating mechanism state evaluation method.
Detailed Description
The present invention will be described in further detail with reference to specific embodiments in order to make the technical field better understand the scheme of the present invention. It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the invention. All other embodiments, which can be obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.
The present invention is described in further detail below with reference to the attached drawing figures.
The invention provides a shore bridge running mechanism state evaluation method based on deep learning, which promotes the health state of a shore bridge key running mechanism to carry out intelligent evaluation, firstly collects vibration signals of the key running mechanism in a shore bridge machine room, calculates the relevant characteristic indexes of the collected vibration signals, then adaptively extracts signal state characteristic indexes based on a built 1D convolutional neural network model, performs weighted fusion on the characteristic indexes, outputs results in a grading mode, can evaluate the state of a shore bridge according to the size and stability of a score, and has the following implementation steps as shown in figure 1:
s1, signal acquisition: in the embodiment, 13 vibration acceleration sensors are used for acquiring data of a lifting motor, a lifting mechanism reduction box and a base in a shore bridge machine room, vibration signals are measured at the same measuring point respectively in 3 mutually orthogonal directions, the vibration overall appearance of the lifting mechanism can be basically reflected, and the 2560HZ sampling rate is set.
Step S2, calculating the characteristic index of the vibration signal by using a formula in the invention content, wherein the method comprises the following steps: calculating the characteristic indexes once every 5s according to the effective value, the average value, the standard deviation, the vibration intensity, the kurtosis, the skewness and the frequency spectrum, and taking the calculation result as the input of the convolutional neural network.
S3, training a model: based on a deep learning algorithm, a 1D Convolutional Neural Network (CNN) is built, collected tagged quay-bridge data is used as a training set, step S2, data vibration characteristic indexes are calculated and input as convolutional Neural Network models, the characteristic indexes are learned, and the convolutional Neural Network models are trained.
Preferably, for a newly-used faultless shore bridge, the shore bridge is just produced and used and is detected to be faultless, the health state of the shore bridge can be considered to be the best, the score is 100, and the collected data is manually labeled with a label 100; for the abnormally maintained shore bridge, because the working environment of the shore bridge is poor, the shore bridge is operated under heavy load for a long time, and the like, the key mechanism of the shore bridge is easy to break down after being operated for a period of time, and the shore bridge is not maintained normally, so that the health state of the shore bridge is considered to be poor or even the shore bridge cannot work, the grade is 0, and the label 0 is manually pasted on the collected data.
Preferably, the convolutional neural network model in the step S3 is trained to build a 1D convolutional neural network, which includes a convolutional layer, a pooling layer, and a fully-connected layer, and neurons in adjacent layers are connected in different ways to realize layer-by-layer transmission of input information; in this example, the network has 8 layers, of which the first 5 convolutional layers (Conv 1 to Conv 5) and the last three layers are all connection layers (FC 1, FC2, FC 3). The first five layers are a feature extraction optimization part, the second two layers are regression models, and the neurons of the adjacent layers are connected in different modes to realize the layer-by-layer transmission of input sample information, and the structure is shown in fig. 2.
Firstly, setting related parameters of a convolution kernel and a pooling layer:
convolutional layer Conv1:32 × 1 × 8 (convolution kernel width × stride passage number);
maximum pooling layer Pool1:2 × 1 × 8 (convolution kernel width × stride channels number);
convolutional layer Conv2:16 × 1 × 16 (convolution kernel width × stride passage number);
maximum pooling layer Pool2:2 × 1 × 16 (convolution kernel width × stride channels);
convolutional layer Conv3:8 × 1 × 32 (convolution kernel width × stride channels number);
maximum pooling layer Pool3:2 × 1 × 32 (convolution kernel width × stride channels);
convolutional layer Conv4:8 × 1 × 32 (convolution kernel width × stride passage number);
maximum pooling layer Pool4:2 × 1 × 32 (convolution kernel width × stride channels);
convolutional layer Conv5:3 × 1 × 64 (convolution kernel width × stride passage number);
maximum pooling layer Pool5:2 × 1 × 64 (convolution kernel width × stride channels).
And further, inputting the calculated characteristic indexes in a matrix form, sending the characteristic indexes into a convolutional layer, performing left-to-right weighted sum on convolutional kernels in the convolutional neural network, setting the size of the convolutional kernels, inputting each region of the matrix by a convolutional kernel sliding window during convolutional operation, multiplying the convolution kernel sliding window by corresponding elements of the region after overturning, accumulating, learning parameters in the convolutional kernels, extracting bottom layer characteristics, and sending the bottom layer characteristics into a pooling layer.
Further, the feature dimensionality is still large after convolution, pooling operation is needed, the embodiment selects the largest pooling layer to perform pooling operation, namely, the largest value is found in each region, secondary feature extraction is performed while the data dimensionality is reduced, and the largest pooling layer can effectively reduce the deviation of the estimated mean value caused by parameter errors of the convolution layer relative to other types of pooling layers.
Further, after convolution and pooling of 5 layers of convolution layers, the bottom layer characteristics are converted into more abstract high layer characteristics layer by layer, and the results are output to a full connection layer.
Preferably, in this example, a ReLU activation function is used, and the activation function introduces a nonlinear factor to a neuron, so that the neural network can approach any nonlinear function at will, when the input is a positive number, the problem of gradient saturation does not exist, and since the score is necessarily greater than or equal to 0, the problem of function non-activation does not occur.
Setting relevant parameters of the full connection layer: full-connectivity layer parameter 82 x 50 x 1 (number of input layer neurons, number of hidden layer neurons, number of output layer neurons).
Further, the full connection layer (FC 1) is used for connecting the output of the maximum pooling layer Pool5, the full connection layer integrates the features extracted by the convolution layer and the pooling layer, the features learned by the network are mapped into the mark space of the sample, the last full connection layer (FC 3) outputs an evaluation value, a Sigmoid activation function is adopted, and the function output value is between 0 and 100, namely the state of the corresponding quay bridge running mechanism is scored.
Further, outputting the obtained result, calculating errors among sample labels, calculating the errors according to the MSE loss function in the invention content, and transmitting the errors from the output to the input layer by layer for back propagation. Firstly, each parameter
And b (i) Initializing a small, near zero random value and then using the gradient descent ruleAnd updating the weight of each characteristic index, training the parameters by data for multiple times, continuously updating the weight and deviation, finding W and b of the minimized loss function, minimizing the error, and finishing the training when the set working state identification accuracy is reached.
Preferably, an MSE loss function formula is selected in the column, the quality of the prediction capability of the model is measured by calculating the mean square error between the sample label and the result obtained by outputting, and compared with other loss functions, the loss function is generally used for a regression model, and the output is a predicted value, which is suitable for the present example.
Preferably, in this example, the weight initialization uses the He initialization method in conjunction with the ReLU activation function, and the He initialization method makes the input and output follow the same distribution as much as possible, so as to avoid that the output value of the activation function of the back layer tends to 0.
Further, the Dropout technique and L2 regularization are employed in this example:
preferably, the Dropout technology randomly deactivates the neurons with a certain probability in the training process, so that the weights of some nodes of the network do not work at random, in this example, the weights are uniformly distributed, 50% of the weights of the nodes are randomly selected to pause for one time and do not work for each input data, so that model averaging is realized, and the generalization capability of the network is improved.
Preferably, the L2 regularization adds the sum of squares of the weighting parameters to the original loss function, in order to limit the parameters from being too large or too large, so that the model is more complex and overfitting is avoided.
Further, experimental result analysis shows that after training of the training set, the training model with good performance and capable of representing the health state of the shore bridge operating mechanism is finally obtained.
Step S4, state evaluation: in the embodiment, the state of the shore bridge running mechanism is analyzed by grading the real-time vibration signals acquired by the hoisting motor in the shore bridge machine room by using the model established in the step S3 according to the data processing mode in the step S2, and the analysis of experimental results shows that the grade can accurately express the health state of the shore bridge running mechanism.
According to the embodiment, the state evaluation method of the shore bridge running mechanism is based on deep learning. By utilizing the excellent automatic characteristic index extraction and nonlinear mapping functions of the convolutional neural network, the method does not need expert experience foundation, does not need manual selection and design of key characteristic indexes, adaptively extracts signal state characteristic indexes, performs characteristic fusion and effectively improves the recognition rate of the state of the shore bridge operating mechanism. And outputting the result in a scoring mode, and visually displaying the health state of the shore bridge operating mechanism according to the scoring height and the scoring stability. Therefore, the method provided by the invention can effectively evaluate the health state of the shore bridge running mechanism, and is beneficial to the development of port machinery towards the direction of automation, unmanned and intellectualization.
Claims (5)
1. A shore bridge running mechanism state evaluation method based on deep learning is characterized by comprising the following steps:
step S1: collecting vibration signals of a key operating mechanism in a shore bridge machine room by using an acceleration sensor;
step S2: calculating a characteristic index commonly used by a vibration signal of a shore bridge running mechanism;
and step S3: selecting labeled data as a training set based on a deep learning algorithm, constructing a training 1D convolutional neural network model, taking a data characteristic index obtained by calculation as input, calculating an output value of each neuron in a forward direction, calculating an error with an actual value, then performing back propagation, calculating a gradient of a connection weight of each neuron, updating the weight of each characteristic index according to a gradient descent rule, finishing training when a set working state identification accuracy is reached, and obtaining a model with a grading estimation capability on the health state of a shore bridge operating mechanism; the method is characterized in that a 1D convolutional neural network model is built, the convolutional neural network comprises a convolutional layer, a pooling layer and a full-connection layer, the neuron output of the last full-connection layer is a shore bridge state score, a Sigmoid activation function is used, and the formula is as follows:
wherein, W i Is a convolution kernel weight matrix;
is a characteristic index of the l-th layer; t is a Sigmoid function input value, a function output value S (t) is between 0 and 100, and a result is converted into a score output from 0 to 100 through a Sigmoid activation function of linear regression;
and step S4: and inputting real-time data to obtain the evaluation of the operating health state of the shore bridge based on the construction of a trained 1D convolutional neural network model, and outputting a result in a scoring mode.
2. The state evaluation method of the shore bridge running mechanism based on deep learning as claimed in claim 1, wherein: in the step S1, the key operation mechanism in the shore bridge machine room comprises a lifting mechanism, a pitching mechanism and a trolley operation mechanism, the lifting mechanism influences the normal operation of the shore bridge system, and the vibration signal of the shore bridge key operation mechanism is acquired by an acceleration sensor for evaluation.
3. The state evaluation method for the shore bridge running mechanism based on deep learning as claimed in claim 1, wherein: and S2, calculating common characteristic indexes of the vibration signals of the shore bridge mechanism, wherein the common characteristic indexes comprise effective values, mean values, standard deviations, vibration intensity, kurtosis, skewness and frequency spectrums.
4. The state evaluation method of the shore bridge running mechanism based on deep learning as claimed in claim 1, wherein: and S4, outputting the current working state of the key operation mechanism in the bank bridge machine room by using the model built in the S3 according to the acquired real-time vibration signal of the key operation mechanism in the bank bridge machine room in the data processing mode of the S2.
5. The state evaluation method for the shore bridge running mechanism based on deep learning as claimed in claim 4, wherein: and (4) adopting a new scoring type intelligent evaluation standard, outputting a result with a score of 0-100, quantizing the evaluation standard, and knowing the health state of the shore bridge operating mechanism according to the output score.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910235249.3A CN109977869B (en) | 2019-03-27 | 2019-03-27 | Shore bridge running mechanism state evaluation method based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910235249.3A CN109977869B (en) | 2019-03-27 | 2019-03-27 | Shore bridge running mechanism state evaluation method based on deep learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109977869A CN109977869A (en) | 2019-07-05 |
| CN109977869B true CN109977869B (en) | 2023-04-07 |
Family
ID=67080764
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910235249.3A Active CN109977869B (en) | 2019-03-27 | 2019-03-27 | Shore bridge running mechanism state evaluation method based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109977869B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110472753B (en) * | 2019-08-22 | 2023-04-18 | 苏宝炜 | Equipment facility unit evaluation method and device based on deep learning |
| CN111783854B (en) * | 2020-06-18 | 2022-06-07 | 武汉互创联合科技有限公司 | Intelligent embryo pregnancy state prediction method and system |
| CN112132430B (en) * | 2020-09-14 | 2022-09-27 | 国网山东省电力公司电力科学研究院 | Reliability evaluation method and system for distributed state sensor of power distribution main equipment |
| CN113570478B (en) * | 2021-06-29 | 2023-10-31 | 中北大学 | Bridge health state intelligent evaluation method based on edge calculation |
| CN114781875B (en) * | 2022-04-25 | 2024-08-02 | 四川中电启明星信息技术有限公司 | Micro-grid economic operation state evaluation method based on deep convolution network |
| CN116467579B (en) * | 2023-04-12 | 2023-11-14 | 宁波送变电建设有限公司运维分公司 | Power equipment health grading method and system based on feature mining technology |
| CN116957366B (en) * | 2023-09-20 | 2024-01-16 | 中交第四航务工程勘察设计院有限公司 | Port data acquisition and evaluation method and system based on Internet of things |
| CN117236802A (en) * | 2023-11-15 | 2023-12-15 | 青岛民航凯亚系统集成有限公司 | Civil airport operation navigability assessment method and system |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108154190A (en) * | 2018-01-12 | 2018-06-12 | 上海海事大学 | A kind of gantry crane motor status clustering method based on SOM neural networks |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108932480B (en) * | 2018-06-08 | 2022-03-15 | 电子科技大学 | Distributed optical fiber sensing signal feature learning and classifying method based on 1D-CNN |
-
2019
- 2019-03-27 CN CN201910235249.3A patent/CN109977869B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108154190A (en) * | 2018-01-12 | 2018-06-12 | 上海海事大学 | A kind of gantry crane motor status clustering method based on SOM neural networks |
Non-Patent Citations (1)
| Title |
|---|
| 基于SOM神经网络的岸桥起升电机状态聚类分析;唐刚等;《东华大学学报(自然科学版)》;20180825(第04期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109977869A (en) | 2019-07-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109977869B (en) | Shore bridge running mechanism state evaluation method based on deep learning | |
| CN111967486A (en) | Complex equipment fault diagnosis method based on multi-sensor fusion | |
| CN109612708B (en) | Power transformer on-line detection system and method based on improved convolutional neural network | |
| CN105300692B (en) | A kind of bearing failure diagnosis and Forecasting Methodology based on expanded Kalman filtration algorithm | |
| CN111830408A (en) | Motor fault diagnosis system and method based on edge calculation and deep learning | |
| She et al. | Wear indicator construction of rolling bearings based on multi-channel deep convolutional neural network with exponentially decaying learning rate | |
| CN109407654B (en) | Industrial data nonlinear causal analysis method based on sparse deep neural network | |
| CN104598993B (en) | Driver tired driving Forecasting Methodology based on number of winks | |
| CN111461187B (en) | Intelligent building settlement detection system | |
| CN101441728A (en) | Neural network method of crane optimum design | |
| CN111310686B (en) | Steel wire rope broken wire damage identification method of IPSO-BP neural network | |
| CN110737976A (en) | mechanical equipment health assessment method based on multi-dimensional information fusion | |
| CN109934422A (en) | Neural network wind speed prediction method based on time series data analysis | |
| CN111738478A (en) | Wave compensation prediction method based on random forest algorithm and Adam neural network | |
| CN104085789B (en) | A kind of intelligent monitoring method of crane running status | |
| CN116186633A (en) | Power consumption abnormality diagnosis method and system based on small sample learning | |
| CN102707713A (en) | Fault diagnosis system and method for automobile safety air bag assembly working procedure | |
| CN113449912B (en) | Space load situation sensing method based on artificial intelligence technology | |
| CN112070103A (en) | Method for inverting atmospheric visibility through microwave link network gridding self-adaptive variable scale | |
| CN113126489B (en) | Intelligent BIT design method of heavy-duty gas turbine control system based on CNN-GRU-BINN | |
| CN117390414A (en) | Intelligent recognition and positioning method and system for water conservancy emergency irrigation and drainage equipment | |
| CN113627268A (en) | Model training method, and method and device for detecting fault of speed reducer for mine hoist | |
| CN110827134A (en) | Power grid enterprise financial health diagnosis method | |
| CN114720129B (en) | Rolling bearing residual life prediction method and system based on bidirectional GRU | |
| CN116662925A (en) | Industrial process soft measurement method based on weighted sparse neural network |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |