CN113987905A - Escalator braking force intelligent diagnosis system based on deep belief network - Google Patents
Escalator braking force intelligent diagnosis system based on deep belief network Download PDFInfo
- Publication number
- CN113987905A CN113987905A CN202011506490.4A CN202011506490A CN113987905A CN 113987905 A CN113987905 A CN 113987905A CN 202011506490 A CN202011506490 A CN 202011506490A CN 113987905 A CN113987905 A CN 113987905A
- Authority
- CN
- China
- Prior art keywords
- escalator
- braking
- braking force
- brake
- data
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- 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
- G06N3/045—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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/048—Activation functions
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/088—Non-supervised learning, e.g. competitive learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N7/00—Computing arrangements based on specific mathematical models
- G06N7/01—Probabilistic graphical models, e.g. probabilistic networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/08—Probabilistic or stochastic CAD
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/28—Fuselage, exterior or interior
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/02—Reliability analysis or reliability optimisation; Failure analysis, e.g. worst case scenario performance, failure mode and effects analysis [FMEA]
Abstract
The invention discloses an escalator braking force intelligent diagnosis system based on a deep belief network, which comprises a braking distance acquisition module, a braking amplitude acquisition module, a braking performance database and a monitoring module, wherein the braking distance acquisition module acquires braking distance, braking amplitude and escalator technical parameters and uploads the data to the braking performance database; B. building an escalator braking force prediction model and building an escalator braking force fault diagnosis model through deep learning; C. the brake performance database stores and reduces the data; D. comparing the reduced data with a braking force prediction model and a braking force fault diagnosis model; E. and checking real-time data and early warning through the braking force intelligent diagnosis platform according to the comparison result. The escalator brake force early warning method has the advantages that the brake force range of the escalator is estimated by using the deep belief network, useful fault characteristics are directly extracted from escalator fault original data, fault states are identified, early warning of brake force abnormity and diagnosis results of brake force faults can be obtained in the first time, hidden dangers existing in the escalator are timely checked, intelligent management is achieved, and the operation efficiency of the escalator is improved.
Description
Technical Field
The invention relates to an escalator braking force detection system; in particular to an escalator braking force intelligent diagnosis system based on a deep belief network.
Background
The escalator and the moving sidewalk are special equipment for continuously transporting personnel, become important personnel transportation tools in public places due to the characteristics of stable operation, safety, reliability, strong conveying capacity, convenient use and the like, and are widely applied to public places such as airports, stations, markets, supermarkets, hospitals and the like. Meanwhile, due to the reasons of large using amount, frequent use, large passenger capacity, complex use environment and the like, accidents of the escalator sometimes occur, and the accidents often have the characteristics of more personnel involved, high injury degree and the like, so that how to ensure the safe operation of the escalator becomes a new research direction.
The investigation and analysis of the accident cases of the escalator and the moving sidewalk show that the following reasons are mainly caused, and the accident reason and the accident proportion are shown in table 1.
The reversion accidents of the escalator are not the most frequent, but the reversion accidents of the escalator have the characteristics of strong group property and social reverberation, and generally occur in the escalator running fully. The mechanical failures that cause the reversal of the escalator are: the driving chain is broken, the step chain is broken and the braking torque of the working brake is insufficient. Because the safety margin is fully considered during the design of the driving chain and the step chain, the probability of breakage is low, the brake wheel and the brake shoe of the escalator generate friction when the escalator is stopped, the brake shoe is abraded, if the escalator is not maintained properly, the brake force is insufficient, and the reversion risk is greatly improved. At present, in the maintenance and regular inspection of an escalator, the braking capacity of the escalator is usually verified indirectly by measuring the braking distance of the escalator, but the braking distance comprises the measurement results under two working conditions of load and no load, the regular inspection only carries out no-load test and supervision and inspection to carry out load test, the inspection frequency is not high, and a lot of potential safety hazards cannot be found in time in the inspection. Factors influencing the braking distance of the escalator include lubrication conditions of moving parts such as step chains and chain wheels, braking force performance and the like, and the condition that whether the braking force is qualified or not is judged from the braking distance singly is not enough.
The existing escalator on-line detection system can monitor and remotely know the service condition and the operation rule of the escalator in real time; the escalator brake force early warning system has certain escalator brake force early warning capacity, can not find out the regular potential risk in time, has risk of risk enlargement, and does not have intelligent diagnosis on the escalator brake force. Under the great trend of the current industry of 'internet' + 'special equipment detection', the research of the on-line monitoring and intelligent early warning technology of the special equipment gradually tends to be perfect in the elevator field. Therefore, the research on the intelligent diagnosis system for the braking force of the escalator is particularly important.
Disclosure of Invention
The invention aims to solve the technical problem of providing an escalator braking force intelligent diagnosis system which can identify the abnormal change condition of the braking force, provide technical support for maintenance and repair of an escalator for maintenance personnel, integrate real-time monitoring, data statistics, intelligent early warning and fault diagnosis and improve the running efficiency of the escalator.
The invention adopts the technical scheme that the escalator braking force intelligent diagnosis system based on the deep belief network comprises the steps of,
A. collecting the braking distance, the brake amplitude and the technical parameters of the escalator and uploading the data to a braking performance database;
B. building an escalator braking force prediction model and building an escalator braking force fault diagnosis model through deep learning;
C. the brake performance database stores and reduces the data;
D. comparing the reduced data with a braking force prediction model and a braking force fault diagnosis model;
E. and checking real-time data and early warning through the braking force intelligent diagnosis platform according to the comparison result.
The step A, acquiring a braking distance, wherein the acquisition comprises an encoder module and a braking signal module which are respectively connected with a processor; the processor wirelessly transmits the processed braking distance to a braking performance database;
the acquisition of the brake amplitude comprises a vibration sensor connected with a processor, and the processor wirelessly transmits the processed brake amplitude data to a brake performance database.
B, deeply learning and establishing an escalator braking force estimation model comprising two layers of RBMs and one layer of BP; the two layers of RBMs are in full connection, and the BP network is in unidirectional connection;
the first layer of RBM display layer is used as a braking force influence factor and comprises the lifting height of the escalator, the nominal speed, the nominal width, the inclination angle, the amplitude of a braking arm, the maximum value of an initial braking force, the no-load up braking distance, the no-load down braking distance and the use environment grade, and 9 neurons of the input layer are input;
drawing up 90 neurons of the RBM of the second layer;
the second hidden layer h2 is an input layer of the BP network, and finally, a power prediction result is output through an output layer of the BP network.
The method comprises the steps of establishing an escalator brake force fault diagnosis model, directly extracting useful fault characteristics from original fault data and identifying fault states;
defining the type of the brake force fault of the escalator;
acquiring braking distance data and brake vibration data of the type of the braking force fault of the escalator as input layer parameters;
adding a classifier on the top layer of the input layer, inputting the preprocessed fault data into a DBM braking force early warning model, and extracting the characteristics of fault diagnosis characterization data from the model layer by layer;
and inputting the test set into a trained diagnostic model to identify the health condition of the equipment.
The escalator brake force fault types comprise brake jamming, asynchronous brake arm actions and insufficient brake torque.
And multi-sensor information processing for preprocessing data is further included between the brake performance database and the intelligent brake force diagnosis platform, and the multi-sensor information processing is a CPCA _ DTW multi-dimensional time sequence.
The invention has the advantages that the braking force range of the escalator is estimated by utilizing the deep belief network, useful fault characteristics are directly extracted from the original data of the escalator faults, and the fault state is identified, so that early warning of braking force abnormity and diagnosis results of the braking force faults can be obtained at the first time, hidden dangers of the escalator are timely checked, intelligent management is realized, and the operation efficiency of the escalator is improved.
Drawings
FIG. 1 is a schematic diagram of a Restricted Boltzmann Machine (RBM) model architecture;
FIG. 2 is a schematic diagram of a Deep Belief Network (DBN) model structure;
FIG. 3 is a general block diagram of the present invention;
FIG. 4 is a block diagram of an escalator stopping distance on-line detection device;
FIG. 5 is a block diagram of an escalator brake amplitude on-line detection device;
FIG. 6 is a schematic structural diagram of an escalator braking force prediction model established based on a DBN;
FIG. 7 is a schematic diagram illustrating a process of fine-tuning network weights by the BP algorithm;
FIG. 8 is a process schematic of DBN-based fault diagnosis model building and training;
FIG. 9 is a schematic diagram of a data reduction process based on CPCA _ DTW multidimensional time series;
fig. 10 is a working flow chart of the intelligent diagnosis monitoring platform for the braking force of the escalator.
Detailed Description
Factors influencing the braking force of the escalator are many, no specific mathematical model is provided for researching the relation between the braking force and the braking distance and other factors, and different models have certain difference under different use environments, so that the braking force model of each escalator is unique. Along with the popularization of leading-edge science and technology, the application of the artificial intelligence machine learning method in special equipment is more and more extensive. A Deep Belief Network (DBN) of unsupervised learning in Deep learning is derived from the research of a simulated brain nervous system of bionics, and the learning process of a human brain is simulated through mutual activation between a apparent layer neuron and a hidden layer neuron, so that the cognition and judgment of an object from original data are directly realized. The deep belief network theory can be applied to the prediction of the escalator braking force value and the diagnosis of braking force abnormity, the basic parameters, the braking distance and the braking arm amplitude of the escalator are classified and identified by directly starting from original data, and the artificial feature extraction process is not needed, so that the human participation factors are reduced, and the prediction accuracy of the escalator braking force value and the intelligence of the braking force abnormity judgment are enhanced.
As shown in fig. 1, a Restricted Boltzmann Machine (RBM) model is a stochastic neural network model having a two-layer structure of a visible layer and a hidden layer, which can be understood as a special markov random field, and has a feature that neurons in the visible layer and the hidden layer are not connected, and neurons between the visible layer and the hidden layer are all connected.
In the RBM, v represents a apparent layer neuron, h represents a hidden layer neuron, W represents the weight of the connection strength between any two adjacent neurons, a represents the apparent layer neuron self-bias, and b represents the hidden layer neuron self-bias. The energy of an RBM can be expressed as:
the probability of the hidden layer neuron hj being activated is as follows:
P(hj|v)=σ(bi+∑iWijvi) (2)
because the hidden layer and the display layer are fully connected, the display layer can be activated by the hidden layer;
P(vi|h)=σ(ai+∑jWijhj) (3)
wherein the content of the first and second substances,is sigmoid function, when the function value field is [0,1 ]]To describe the probability of neurons being activated. The neurons in the same layer are not connected, so the neurons satisfy the independence of probability surface density:
after a certain piece of data x is given to the display layer, the RBM can calculate the probability P (h) of each hidden layer neuron being activated according to the formula (2)j| x), take μ ∈ [0,1 ]]As the threshold, when P (h)j| x) ≥ mu hjWhen P (h) is equal to 1j|x)<μ hour hjFrom this, it can be derived whether each hidden layer nerve is activated. An RBM model is determined based on three parameters, a, b, and W.
As shown in fig. 2, a Deep Belief Network (DBN) is composed of a plurality of layers of neurons, a component of the Deep Belief Network is a Restricted Boltzmann Machine (RBM), the plurality of RBMs are connected in series to form a DBN, and a hidden layer of a previous RBM is a display layer of a next RBM, as shown in fig. 2. Therefore, the deep belief network can also be interpreted as a Bayesian probability generation model consisting of multiple layers of random hidden variables. The first layer of the display layer can be input data, in the training stage, relevant information is extracted from the display layer through Gibbs sampling and mapped to the hidden layer, information is extracted from the hidden layer through the Gibbs sampling and mapped to the display layer, the input data is reconstructed in the display layer, the mapping and reconstruction processes between the display layer and the hidden layer are repeatedly executed, and the values of the weights a, b and W are continuously updated in each reconstruction process.
The invention is described in further detail below with reference to the following figures and detailed description:
as shown in fig. 3, the escalator braking force intelligent diagnosis system based on the deep belief network of the present invention comprises the steps of,
A. collecting the braking distance, the brake amplitude and the technical parameters of the escalator and uploading the data to a braking performance database;
as shown in fig. 4, the acquisition of the braking distance includes an encoder module and a braking signal module respectively connected with the processor; the encoder module is installed and is connected with the treater through the cable at handrail area return section entrance, and the braking signal module is installed and is connected with the treater through the cable on switch board band-type brake contactor, and the treater will handle the braking distance wireless transmission who reachs to the braking performance database.
As illustrated in fig. 5, the acquisition of brake amplitude includes a vibration sensor connected to a processor; the vibration sensors are installed on the two brake arms, when the escalator brake is opened and closed every time, the vibration sensors are triggered to work, the vibration sensors are connected with the processor, and the processor wirelessly transmits processed brake amplitude data to the brake performance database.
B. The braking force prediction model of the escalator and the braking force fault diagnosis model of the escalator are established through deep learning, and the braking force ranges of the escalators in different models and different use environments can be predicted;
as shown in fig. 6, the model for estimating braking force of the escalator through deep learning and building comprises two layers of RBMs and one layer of BP; the two layers of RBMs are in full connection, and the BP network is in unidirectional connection;
the first layer of RBM display layer is used as a braking force influence factor and comprises the lifting height of the escalator, the nominal speed, the nominal width, the inclination angle, the amplitude of a braking arm, the maximum value of an initial braking force, the no-load up braking distance, the no-load down braking distance and the use environment grade, and 9 neurons of the input layer are input;
drawing up 90 neurons of the RBM of the second layer; the second hidden layer h2 is an input layer of the BP network, and finally, a power prediction result is output through an output layer of the BP network.
As shown in fig. 7, the deep belief network training process is: directly taking basic technical parameters of equipment as input layer parameters; and carrying out data preprocessing on the braking distance signal acquired in real time and the vibration signal of the brake arm, and taking the processed data as an input layer parameter. Determining a learning rate, a learning direction, a maximum iteration number of training, a maximum iteration number of fine tuning, a sample number and a training batch, setting a Neural Network (Neural Network) NN threshold function as a Sigmoid function, and calculating the probability of h1 through v1, Gibbs sampling: h is1~P(h|v1) Then, the display layer v1 is reconstructed through h1, namely, the probability distribution P (v | h) of the display layer is reversely deduced by utilizing the hidden layer1) And after the accuracy requirement is met, the first layer of RBM training is finished. Second tier RBM training begins, h1 as the presentation tier for the second tier RBMv2 as input, continuing the RBM training method described above, calculating the probability of v2 by h1, Gibbs sampling: v. of2~P(v|h1) Calculating the probability gibbs sampling h of h2 by v22~P(h|v2) And finishing the training of the RBM of the second layer after the requirements are met. The training result of h2 is supervised training for the input factor of BP network, fine tuning the network weight according to BP error back propagation algorithm, updating the weight content:
W←W+λ(P(h1|v1)v1-P(h2|v2)v2)
a←a+λ(v1-v2)
b←b+λ(v1-v2)
therefore, the model converges to a local optimal point, and finally, the estimated braking force value is obtained, and the whole estimated model network is completed.
In the aspect of fault diagnosis, when the deep belief network faces complex and huge data, the defect of insufficient feature expression capability of the traditional shallow learning method can be effectively overcome, and useful fault features can be directly extracted from the original data of the fault and the fault state can be identified.
As shown in fig. 8, the building and training of the DBN-based escalator fault diagnosis model:
1) the method is characterized in that the fault type of the braking force of the escalator is determined firstly, and the faults can be basically divided into brake blocking, asynchronous brake arm action, insufficient braking torque and the like.
2) And (3) artificially simulating the states, and acquiring braking distance data and brake vibration data in the corresponding states as input layer parameters.
3) And adding a classifier on the top layer of the input layer, inputting the preprocessed fault data into a DBM braking force early warning model, training the model by utilizing a fault sample, and extracting the characteristics of the fault diagnosis characterization data from the model layer by layer.
4) And (3) carrying out supervised fine adjustment on related parameters of the DBN by using a network, and inputting the test set into a trained diagnosis model after the fine adjustment is finished so as to identify the health condition of the equipment.
C. The brake performance database stores and reduces the data;
the parameters of braking distance signals and vibration signals of the brake arm are collected, the two signals are collected respectively by means of a displacement sensor and a vibration sensor, and after the signals are obtained, data are preprocessed to be used as input layer parameters to conduct fault diagnosis. The data preprocessing is an information processing mode of performing integration and reduction on data of two different sensor signals to obtain multi-sensor characterization data.
Fault diagnosis methods based on multi-sensor monitoring information processing can be roughly classified into four categories: data layer fusion, feature layer fusion, decision layer fusion and multi-dimensional time series classification. The invention adopts a multi-dimensional time series classification method.
The multi-dimensional time series classification methods mainly include a distance-based method and a feature-based method. The two methods are combined to form the CPCA _ DTW-based multidimensional Time sequence, namely, based on Common Principal Component Analysis (CPCA) and Dynamic Time Warping (DTW). The CPCA reduces the dimension of the sequence on the basis of ensuring the extraction of the features, and the DTW distance can search the minimum distance metric value between two time sequences by a certain optimization method.
D. Comparing the reduced data with a braking force prediction model and a braking force fault diagnosis model;
as shown in fig. 9, the specific scheme is to reduce the dimensionality of the signal sample by using CPCA, calculate an average covariance matrix and eigenvectors, and sort the eigenvalues to obtain a reduced sequence; calculating the distance between every two vectors in the reduced sequence by using the DTW distance, recording a distance matrix, and stacking the reduced sequence to form a one-dimensional vector; and finally, combining the matrix and the one-dimensional vector to form a one-dimensional data vector which can be input into the DBN model.
E. And checking real-time data and early warning through the braking force intelligent diagnosis platform according to the comparison result.
As shown in fig. 10, by using an escalator braking force estimation model based on a deep belief network, we can estimate the range of an escalator braking force; by the escalator fault diagnosis model based on the deep belief network, diagnosis of braking force related faults is achieved. On the basis, the intelligent braking force diagnosis monitoring platform is built to realize informatization monitoring, and real-time monitoring, data statistics, intelligent early warning and fault diagnosis are integrated.
The arrangement of the acquisition equipment is completed on the escalator site, and the acquisition equipment comprises a braking distance detection sensor and a brake vibration monitoring sensor. And establishing and training a braking force prediction model and a braking force fault diagnosis model in a cloud system. When the escalator is normally used, the acquisition device is used for extracting the data related to braking force under the working condition of normal stop or emergency braking at each time, namely the amplitude of the brake, the ascending/descending braking distance and the like, and the data are stored and reduced in the cloud server of the Internet of things. And comparing the reduced data with the pre-estimation model and the fault diagnosis model to realize monitoring analysis, entering an early warning range when the braking force is reduced to a sub-health state, and sending an early warning signal to an Internet of things cloud server to remind a user unit and a maintenance unit to take corresponding measures. When a suspected fault occurs, the system extracts fault characteristics, compares the fault characteristics with a fault diagnosis model, diagnoses the brake force fault in real time, and comprehensively analyzes the brake force state through the estimated value given by the estimation model and the fault type diagnosed by the fault diagnosis model. After the data is acquired, the user can view the real-time data, the historical data and the diagnosis result through the browser.
The using unit can master the braking force condition of the escalator in real time through the intelligent diagnosis platform, obtain early warning of braking force abnormity and diagnosis results of braking force faults at the first time, and timely troubleshoot hidden dangers of the escalator, so that intelligent management is realized, and the comprehensive service level is improved.
The maintenance unit can timely figure out the reason of the brake force fault through the intelligent diagnosis platform, master the trend of brake force reduction and brake distance increase, pertinently carry out problem troubleshooting, carry out preventive maintenance, reduce the fault rate and accident rate of the escalator, and improve the service quality of the maintenance unit.
The manufacturing unit can carry out long-term information management on the escalator of the factory through the intelligent diagnosis platform, potential factors inducing braking force related faults are mined, data support is provided for the design process of the optimized braking system, and product quality and market competitiveness are improved.
The inspection department can transfer the historical data of the brake force fault of the escalator and the long-term change trend of the brake force through the intelligent diagnosis platform, and the brake force condition and the maintenance quality of the inspected escalator can be conveniently mastered.
It should be noted that the protection scope of the present invention is not limited to the above specific examples, and the object of the present invention can be achieved by substantially the same structure according to the basic technical concept of the present invention, and embodiments that can be imagined by those skilled in the art without creative efforts belong to the protection scope of the present invention.
Claims (6)
1. An escalator braking force intelligent diagnosis system based on a deep belief network is characterized by comprising the following steps,
A. collecting the braking distance, the brake amplitude and the technical parameters of the escalator and uploading the data to a braking performance database;
B. building an escalator braking force prediction model and building an escalator braking force fault diagnosis model through deep learning;
C. the brake performance database stores and reduces the data;
D. comparing the reduced data with a braking force prediction model and a braking force fault diagnosis model;
E. and checking real-time data and early warning through the braking force intelligent diagnosis platform according to the comparison result.
2. The escalator braking force intelligent diagnosis system based on deep belief network of claim 1,
the step A, acquiring a braking distance, wherein the acquisition comprises an encoder module and a braking signal module which are respectively connected with a processor; the processor wirelessly transmits the processed braking distance to a braking performance database;
the acquisition of the brake amplitude comprises a vibration sensor connected with a processor, and the processor wirelessly transmits the processed brake amplitude data to a brake performance database.
3. The escalator braking force intelligent diagnosis system based on deep belief network of claim 1,
b, deeply learning and establishing an escalator braking force estimation model comprising two layers of RBMs and one layer of BP; the two layers of RBMs are in full connection, and the BP network is in unidirectional connection;
the first layer of RBM display layer is used as a braking force influence factor and comprises the lifting height of the escalator, the nominal speed, the nominal width, the inclination angle, the amplitude of a braking arm, the maximum value of an initial braking force, the no-load up braking distance, the no-load down braking distance and the use environment grade, and 9 neurons of the input layer are input;
drawing up 90 neurons of the RBM of the second layer;
the second hidden layer h2 is an input layer of the BP network, and finally, a power prediction result is output through an output layer of the BP network.
4. The escalator braking force intelligent diagnosis system based on deep belief network of claim 1,
the method comprises the steps of establishing an escalator brake force fault diagnosis model, directly extracting useful fault characteristics from original fault data and identifying fault states;
defining the type of the brake force fault of the escalator;
acquiring braking distance data and brake vibration data of the type of the braking force fault of the escalator as input layer parameters;
adding a classifier on the top layer of the input layer, inputting the preprocessed fault data into a DBM braking force early warning model, and extracting the characteristics of fault diagnosis characterization data from the model layer by layer;
and inputting the test set into a trained diagnostic model to identify the health condition of the equipment.
5. The escalator braking force intelligent diagnosis system based on deep belief network of claim 4, characterized in that the escalator braking force failure types include brake jamming, out of synchronization of brake arm actions and insufficient braking torque.
6. The escalator braking force intelligent diagnosis system based on deep belief network of claim 1,
and multi-sensor information processing for preprocessing data is further included between the brake performance database and the intelligent brake force diagnosis platform, and the multi-sensor information processing is a CPCA _ DTW multi-dimensional time sequence.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011506490.4A CN113987905A (en) | 2020-12-18 | 2020-12-18 | Escalator braking force intelligent diagnosis system based on deep belief network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011506490.4A CN113987905A (en) | 2020-12-18 | 2020-12-18 | Escalator braking force intelligent diagnosis system based on deep belief network |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113987905A true CN113987905A (en) | 2022-01-28 |
Family
ID=79731657
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202011506490.4A Pending CN113987905A (en) | 2020-12-18 | 2020-12-18 | Escalator braking force intelligent diagnosis system based on deep belief network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113987905A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116740064A (en) * | 2023-08-14 | 2023-09-12 | 山东奥洛瑞医疗科技有限公司 | Nuclear magnetic resonance tumor region extraction method |
CN114923107B (en) * | 2022-05-05 | 2024-04-05 | 广州广日电梯工业有限公司 | Escalator lubrication control method and device based on Internet of things and cluster analysis |
-
2020
- 2020-12-18 CN CN202011506490.4A patent/CN113987905A/en active Pending
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114923107B (en) * | 2022-05-05 | 2024-04-05 | 广州广日电梯工业有限公司 | Escalator lubrication control method and device based on Internet of things and cluster analysis |
CN116740064A (en) * | 2023-08-14 | 2023-09-12 | 山东奥洛瑞医疗科技有限公司 | Nuclear magnetic resonance tumor region extraction method |
CN116740064B (en) * | 2023-08-14 | 2023-10-20 | 山东奥洛瑞医疗科技有限公司 | Nuclear magnetic resonance tumor region extraction method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US20210374506A1 (en) | Method for predicting bearing life based on hidden markov model and transfer learning | |
Kankar et al. | Rolling element bearing fault diagnosis using wavelet transform | |
CN108569607B (en) | Elevator fault early warning method based on bidirectional gating cyclic neural network | |
CN107941537B (en) | A kind of mechanical equipment health state evaluation method | |
CN111353482B (en) | LSTM-based fatigue factor recessive anomaly detection and fault diagnosis method | |
Wang et al. | Intelligent diagnosis method for rolling element bearing faults using possibility theory and neural network | |
Soualhi et al. | Prognosis of bearing failures using hidden Markov models and the adaptive neuro-fuzzy inference system | |
CN111813084B (en) | Mechanical equipment fault diagnosis method based on deep learning | |
WO2016004774A1 (en) | Rail transportation fault diagnosis method and system based on time series analysis | |
CN113255848B (en) | Water turbine cavitation sound signal identification method based on big data learning | |
CN111505424A (en) | Large experimental device power equipment fault diagnosis method based on deep convolutional neural network | |
CN113762329A (en) | Method and system for constructing state prediction model of large rolling mill | |
CN110405537B (en) | Method for establishing guide rail precision prediction model based on deep learning | |
CN109969895B (en) | Fault prediction method based on elevator operation parameters, terminal and readable storage medium | |
CN113987905A (en) | Escalator braking force intelligent diagnosis system based on deep belief network | |
CN111723449B (en) | Performance degradation evaluation method for constant-deceleration braking system of mine hoist | |
CN110737976B (en) | Mechanical equipment health assessment method based on multidimensional information fusion | |
CN111931849A (en) | Hydroelectric generating set operation data trend early warning method | |
CN112132394B (en) | Power plant circulating water pump predictive state evaluation method and system | |
CN107844067A (en) | A kind of gate of hydropower station on-line condition monitoring control method and monitoring system | |
CN112257914A (en) | Aviation safety cause and effect prediction method based on random forest | |
CN107688820B (en) | Elevator fault diagnosis method based on BCSA optimized support vector machine | |
CN114757365A (en) | High-speed railway roadbed settlement prediction and early warning method based on deep learning | |
CN113158362B (en) | Elevator residual life prediction method integrating physical failure and data driving | |
CN109255201A (en) | A kind of ball screw assembly, health evaluating method based on SOM-MQE |
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 |