CN109948778A - A kind of refrigeration equipment trouble-shooter and system - Google Patents
A kind of refrigeration equipment trouble-shooter and system Download PDFInfo
- Publication number
- CN109948778A CN109948778A CN201910148909.4A CN201910148909A CN109948778A CN 109948778 A CN109948778 A CN 109948778A CN 201910148909 A CN201910148909 A CN 201910148909A CN 109948778 A CN109948778 A CN 109948778A
- Authority
- CN
- China
- Prior art keywords
- neural network
- fault diagnosis
- deep neural
- refrigeration equipment
- shooter
- 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
- 238000005057 refrigeration Methods 0.000 title claims abstract description 63
- 238000003745 diagnosis Methods 0.000 claims abstract description 98
- 238000013528 artificial neural network Methods 0.000 claims abstract description 76
- 238000012549 training Methods 0.000 claims abstract description 43
- 238000005457 optimization Methods 0.000 claims abstract description 20
- 238000003062 neural network model Methods 0.000 claims abstract description 19
- 238000012360 testing method Methods 0.000 claims abstract description 17
- 230000006870 function Effects 0.000 claims abstract description 16
- 238000002922 simulated annealing Methods 0.000 claims abstract description 14
- 230000004913 activation Effects 0.000 claims abstract description 11
- 238000003860 storage Methods 0.000 claims abstract description 10
- 230000015572 biosynthetic process Effects 0.000 claims abstract description 9
- 238000005755 formation reaction Methods 0.000 claims abstract description 9
- 238000011478 gradient descent method Methods 0.000 claims description 7
- 239000011159 matrix material Substances 0.000 claims description 4
- 210000004218 nerve net Anatomy 0.000 claims 1
- 238000000034 method Methods 0.000 description 21
- 238000004891 communication Methods 0.000 description 16
- 238000004422 calculation algorithm Methods 0.000 description 12
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 12
- 230000015556 catabolic process Effects 0.000 description 10
- 230000007423 decrease Effects 0.000 description 9
- 238000010586 diagram Methods 0.000 description 9
- 230000008569 process Effects 0.000 description 9
- 238000000137 annealing Methods 0.000 description 6
- 238000004364 calculation method Methods 0.000 description 6
- 238000002474 experimental method Methods 0.000 description 5
- 238000007726 management method Methods 0.000 description 5
- 230000001537 neural effect Effects 0.000 description 5
- 238000004088 simulation Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000000644 propagated effect Effects 0.000 description 3
- 241000208340 Araliaceae Species 0.000 description 2
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 2
- 235000003140 Panax quinquefolius Nutrition 0.000 description 2
- 230000009471 action Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000002405 diagnostic procedure Methods 0.000 description 2
- 235000013399 edible fruits Nutrition 0.000 description 2
- 235000008434 ginseng Nutrition 0.000 description 2
- 230000017525 heat dissipation Effects 0.000 description 2
- 230000007257 malfunction Effects 0.000 description 2
- 239000003507 refrigerant Substances 0.000 description 2
- 241000272510 Apteryx Species 0.000 description 1
- 241000532370 Atla Species 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 239000010687 lubricating oil Substances 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 238000005272 metallurgy Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
Landscapes
- Air Conditioning Control Device (AREA)
Abstract
Refrigeration equipment trouble-shooter according to the present invention and system, refrigeration equipment trouble-shooter includes management storage unit, node and number of plies configuration part, deep neural network training department, deep neural network model formations, train epochs configuration part, fault diagnosis model building portion, learning rate Optimization Dept. and fault diagnosis result generating unit, management is at least stored with set evidence and test group data with storage unit, deep neural network training department is trained deep neural network using SGDM, deep neural network model formations is according to the structure of neural network, activation primitive, loss function etc. establishes deep neural network model, fault diagnosis model building portion constructs fault diagnosis model after being trained for application training group data in deep neural network model, learning rate Optimization Dept. is using simulated annealing to study Rate optimizes, and fault diagnosis result generating unit carries out fault diagnosis to test group data, obtains fault diagnosis result.
Description
Technical field
The invention belongs to refrigerating field, it is related to a kind of refrigeration equipment trouble-shooter and system, and in particular to Yi Zhongleng
Water dispenser group trouble-shooter and system.
Background technique
The industry of modernization and civil buildings all be unable to do without refrigeration system, and refrigeration system component is complicated, it is caused to run
Frequent Troubles in the process, and it is many kinds of.Increase energy consumption, shortening equipment use the longevity while generation of failure influences refrigeration effect
Security risk is ordered and brings, thereby it is ensured that the running quality of refrigeration system is particularly important.Real-time shape is carried out to refrigeration system
State detection and fault diagnosis, can not only guarantee the normal operation of refrigeration system, but also can find the problem and be tieed up in time
It repairs, in the past few decades, breakdown of refrigeration system diagnosis has been a hot spot of research with detection technique.In addition, neural network is logical
It often has learning rate to be difficult to adjust, initial weight threshold value is difficult to the problems such as determining.
Summary of the invention
The present invention searches out global optimum using what intelligent algorithm optimization can enable deep neural network more efficient.Simulation
Annealing algorithm (Simulated Annealing, SA) is a kind of general optimization algorithm, theoretically with the global optimization of probability
Performance is demonstrated by important and wide development and application space under the historical background of nowadays big data.
SA-DNN (simulated annealing optimization deep neural network) is applied to breakdown of refrigeration system diagnosis, selection by the present invention
SGDM (small lot momentum stochastic gradient descent method) training method, each step gradient calculate one group only randomly selected in training set
Sample improves model convergence rate, in order to reach global optimum, using simulated annealing thought in neural neural network
Habit rate optimizes, and gradually exponentially reduces learning rate, so that error precision is rapidly dropped to a lesser value, and find the overall situation
It is optimal.In consideration of it, the present invention provides a kind of breakdown of refrigeration system diagnostic method.
The present invention provides a kind of refrigeration equipment trouble-shooters, have the feature that, including management storage unit,
Node and number of plies configuration part, deep neural network training department, deep neural network model formations, train epochs configuration part, event
Hinder diagnostic model building portion, learning rate Optimization Dept. and fault diagnosis result generating unit, wherein management is at least stored with storage unit
There are set evidence and test group data, node and number of plies configuration part are used to be arranged the number of nodes and the number of plies of deep neural network,
Deep neural network training department is trained deep neural network using SGDM (small lot momentum stochastic gradient descent method), deep
It spends neural network model formations and deep neural network mould is established according to the structure of neural network, activation primitive, loss function etc.
Type, train epochs configuration part are used to determine the train epochs of deep neural network, and fault diagnosis model building portion is for depth mind
Fault diagnosis model is constructed after application training group data in network model are trained, learning rate Optimization Dept. uses simulated annealing
Algorithm optimizes learning rate, and fault diagnosis result generating unit carries out fault diagnosis to test group data, obtains fault diagnosis
As a result.
In refrigeration equipment trouble-shooter provided by the invention, it can also have the following features: wherein, lose letter
The expression formula of number C are as follows:
In formula, x indicates that sample, y indicate actual value, and a indicates output valve, and n indicates the sum of sample.
In addition, can also have the following features: wherein in refrigeration equipment trouble-shooter provided by the invention,
The output form of diagnostic result can be indicated with the form of confusion matrix.
In addition, can also have the following features: wherein in refrigeration equipment trouble-shooter provided by the invention,
The expression formula of activation primitive σ are as follows:
In formula, y=ω1×x1+ω2×x2+…ωm×xm+b
ωmFor the weight of m-th of hidden layer node of a certain layer, xmFor the input value of m-th of hidden layer node of a certain layer, b is
The threshold value of m-th of hidden layer node of a certain layer.
The present invention provides a kind of refrigeration equipment fault diagnosis system, has the feature that, including data acquisition device；With
And equipment fault diagnosis device, equipment fault diagnosis device are the above-mentioned equipment fault diagnosis device of any one, wherein are set
Management in standby trouble-shooter is stored with the data from data acquisition device acquisition with storage unit.
In refrigeration equipment fault diagnosis system provided by the invention, it can also have the following features: wherein, data are adopted
Acquisition means and equipment fault diagnosis device communicate to connect.
In addition, can also have the following features: wherein in refrigeration equipment fault diagnosis system provided by the invention,
Equipment fault diagnosis device is any one in fixed terminal and mobile terminal.
In addition, can also have the following features: wherein in refrigeration equipment fault diagnosis system provided by the invention,
Fixed terminal includes desktop computer.
In addition, can also have the following features: wherein in refrigeration equipment fault diagnosis system provided by the invention,
Mobile terminal includes smart phone, tablet computer.
The action and effect of invention
Easily sink into local minimum when for common BP neural network optimizing, and network structure is not sufficiently stable, every time training
Obtained result differs the problems such as larger and net training time is too long, in order to improve deep neural network (Deep Neural
Network, DNN) to the efficiency and accuracy of breakdown of refrigeration system diagnosis, refrigeration equipment fault diagnosis dress according to the present invention
It sets and system application small lot momentum stochastic gradient descent (Stochastic Gradient Descent Momentum, SGDM)
Training method is trained deep neural network, and each step gradient calculating only randomly selects one group of sample in training set, improves
Its convergence rate.
In addition, refrigeration equipment trouble-shooter of the invention and system are simultaneously using simulation in order to reach global optimum
Annealing optimization deep neural network (Simulated annealing-Deep Neural Network, SA-DNN) model.As a result
Show: to seven quasi-representative failures of refrigeration system, SA-DNN greatly improves rate of correct diagnosis, and optimal network structure is 2
64 nodes of hidden layer, rate of correct diagnosis 99.3%, time-consuming diagnosis is only 3min50s.(Back is propagated compared to conventional counter
Propagation, BP) neural network, breakdown of refrigeration system diagnostic method of the invention is stable with SA-DNN result, is not easy to fall into
Enter local minimum, more effectively realizes breakdown of refrigeration system diagnosis, and the beneficial effect of function admirable.
Detailed description of the invention
Fig. 1 is refrigeration equipment fault diagnosis system structural block diagram in the embodiment of the present invention；
Fig. 2 is refrigeration equipment trouble-shooter structural block diagram in the embodiment of the present invention；
Fig. 3 is the flow diagram of Simulated Anneal Algorithm Optimize deep neural network in the embodiment of the present invention；
Fig. 4 is DNN deep neural network topological diagram in the embodiment of the present invention；
Fig. 5 is the gradient decline schematic diagram in the embodiment of the present invention without momentum；
Fig. 6 is the schematic diagram that the gradient decline of momentum is added in the embodiment of the present invention；
Fig. 7 is Simulated Anneal Algorithm Optimize learning rate flow diagram in embodiment in the embodiment of the present invention；
Fig. 8 is heating power of the SA-DNN compared with DNN model the number of hidden nodes and rate of correct diagnosis in the embodiment of the present invention
Figure；And
Fig. 9 is SA-DNN, DNN and BP neural network diagnosis performance comparison schematic diagram in the embodiment of the present invention.
Specific embodiment
It is real below in order to be easy to understand the technical means, the creative features, the aims and the efficiencies achieved by the present invention
Example combination attached drawing is applied to be specifically addressed refrigeration equipment trouble-shooter of the invention and system.
Fault Diagnosis of Mechanical Equipment 100 is used for the fault diagnosis of water cooler in embodiment.
Fig. 1 is Fault Diagnosis of Mechanical Equipment structural block diagram in the embodiment of the present invention.
As shown in Figure 1, Fault Diagnosis of Mechanical Equipment 100 include data acquisition device 10, equipment fault diagnosis device 20 with
And communication network 40.
Wherein, the data acquired from data acquisition device 10 are stored in equipment fault diagnosis device 20.
Data acquisition device 10 and equipment fault diagnosis device 20 can take communication connection, and communication connection includes wired company
It connects and is wirelessly connected, the present embodiment is to be wirelessly connected.
Data acquisition device 10 includes multiple acquisition units, data processing unit, control unit and acquisition side communication unit
Member.
Multiple acquisition units are separately positioned in water cooler, for acquiring the operation number of different parts in water cooler
According to.
For the present embodiment by simulation water cooler malfunction test, multiple acquisition units collect multiple fisrt feature ginsengs
Number obtains multiple second feature parameters by calculating on the basis of fisrt feature parameter, merges fisrt feature parameter and second
Characteristic parameter obtains multiple groups third feature parameter.
Breakdown of refrigeration system simulated experiment object used in the present embodiment is a centrifugal refrigerating machines.Use failure mould
Draft experiment platform simulates the experiment of water cooler variety classes, different brackets failure under different operating conditions.It acquires at regular intervals
Data, acquire the characteristic parameter (total a) including temperature, pressure, and control unit control data processing unit is being adopted
Collect by calculating the characteristic parameter (total b is a) including obtaining COP, water flow, heat dissipation capacity on the basis of data, these features
Parameter (number is C=a+b) is combined the feature as characterization water cooler failure.Acquire N group altogether during the experiment
Data (dimension of every group of data is C).Wherein, there are one corresponding tally sets for every group of data, and tally set is the square of a 1xN
Gust, each line number word has respectively represented the state (including whether the type for malfunction and failure) of this group of data in matrix.
Multiple groups third feature parameter is pre-processed, obtain that treated set evidence and test group data.
Influence for removal dimension to model training process, prevents certain one-dimensional or certain apteryx is excessive on data influence, to institute
There are data to be standardized, processing mode is as follows:
1) first: calculating the mean value and standard variance of each feature of N group data (i.e. dimension).Ith feature parameter Xi's
Mean value computation formula are as follows:
Wherein:The mean value of ith feature parameter；
The number of C- characteristic parameter；
The group number of N- acquisition data；
Xj,iThe value of jth group data ith feature parameter.
The standard variance S of ith feature parameteriCalculation formula are as follows:Wherein: Si- the i-th
The standard variance of a characteristic parameter；
The group number of N- acquisition data；
Xj,iThe value of jth group data ith feature parameter；
The mean value of ith feature parameter.
2) all data are standardized after calculating mean value and standard variance, wherein jth group ith feature to
The standardized calculation formula of amount:
Wherein,Value after the standardization of jth group ith feature vector；
Xj,iThe value of jth group data ith feature parameter；
The mean value of ith feature parameter；
SiThe standard variance of ith feature parameter.
It is two groups by all N group data random divisions after the completion of normalized, one group (accounts for about total data for training group
2/3), for SA-DNN (Simulated Anneal Algorithm Optimize deep neural network) model for setting up of training, another group of conduct
Test group (account for about total data 1/3), for testing trained model.
Control unit is connected with multiple acquisition units, acquires the related data from multiple acquisition units and is stored,
Also it may be displayed on the screen of control unit.
Acquisition side communication unit is used to for the related operation data stored in control unit to be sent to equipment fault diagnosis dress
Set 20.In the present embodiment, acquisition side communication unit carries out data transmission by communication network 40 by the way of wireless communication.
As shown in Fig. 2, refrigeration equipment trouble-shooter 20 includes diagnosis side communication unit 21, management storage unit 22, section
Point and number of plies configuration part 23, deep neural network training department 24, deep neural network model formations 25, train epochs configuration part
26, tentative diagnosis model construction portion 27, learning rate Optimization Dept. 28, fault diagnosis result generating unit 29 and the above-mentioned each portion of control
Diagnose side control unit 31.
It diagnoses side communication unit 21 to receive in the related data for acquiring side communication unit, the present embodiment, the communication of diagnosis side
Portion 21 carries out data receiver by communication network 40 by the way of wireless communication.
Management is at least stored with the received correlation from acquisition side communication unit of diagnosis side communication unit 21 with storage unit 22
Data include set evidence and test group data.
The gradual failure detected is not easy for 7 kinds of water cooler and carries out modeling analysis, such as: refrigerant leakage, lubricating oil mistake
Amount, evaporator and condenser water flow deficiency, condenser fouling, refrigerant are containing incoagulable gas etc., and establishing SA-DNN, (simulation is moved back
Fiery algorithm optimization deep neural network) water cooler fault diagnosis model.
Neural network network model by neural network structure (show as the hidden layer number of plies and number of nodes, it is as shown in Figure 4
Network model is five hidden layers, the structural topology figure for the deep neural network that each node in hidden layer is 5), activation primitive
(activation primitive is used to calculate the value of node to next node being connected to, and calculating process uses weight threshold) here
Activation primitive selects Sigmoid, and specific calculation is (with a certain hidden layer, for total m node in hidden layer), activates letter
The expression formula of number σ are as follows:
Y=ω1×x1+ω2×x2+…ωm×xm+b (1)
In formula (1), ωmFor the weight of m-th of hidden layer node of a certain layer, xmFor the defeated of m-th of hidden layer node of a certain layer
Enter value.B is the threshold value of m-th of hidden layer node of a certain layer.
Node and number of plies configuration part 23 are used to be arranged the number of nodes and the number of plies of deep neural network.
Node and number of plies configuration part 23 are first designed the number of nodes of deep neural network and the number of plies, the number of plies from 1,2,
3 gradually test, and number of nodes is set as 2 power (64,128,256,1024 ...).
Deep neural network includes the structure and training two aspects of its method of its more hidden layer.The topological structure of more hidden layers
It is similar with multi-layer perception (MLP), as shown in figure 4, the deep neural network being made of input layer, 5 hidden layers and output layer, passes through
Information is propagated forward, and the method for error back propagation corrects its weight and threshold value.X1, X2 in Fig. 4, X3 ..., X64 be defeated
Enter 64 characteristic parameters, including evaporator and condenser side inlet and outlet temperature, compressor air suction delivery temperature etc. 48 ginsengs
Number for experiment measured by, including 16 parameters such as heat dissipation capacity, water by calculate gained.Y is output fault category, is 0-7, wherein
0 represents the case where cold refrigeration system operates normally, and 1-4 is four local faults, and 5-7 is three system failures, and failure code name is shown in
Table 1.
Table 1
C is loss function, and to avoid the occurrence of gradient dispersion problem, the present embodiment, which is chosen, intersects entropy function (cross
Entropy, CE) it is used as loss function.CE features the distance between two probability distribution, it is in classification problem using comparing
A kind of wide loss function, the formula of the criterion are as follows:
Wherein, x indicates that sample, y indicate actual value, and a indicates output valve, and n indicates the sum of sample.
Deep neural network training department 24 is using neural to depth using SGDM (small lot momentum stochastic gradient descent method)
Network is trained.
Deep neural network training department 24 is using SGDM (small lot momentum stochastic gradient descent method) to deep neural network
It is trained.8000 groups of training samples are divided into 100 groups, every group is trained comprising 80 samples, DNN neural computing
When each step gradient, i.e., every update primary parameter when, one group of sample in training set is randomly selected, in larger data set
On, substantially increase the training effectiveness of network model.And because of its randomness, showed compared with traditional gradient descent method more
Erratic behavior, can easily skip very much local minimum, therefore than the decline calculation of traditional gradient on finding global minimum
Method shows more advantages.
In stochastic gradient descent training method, momentum is introduced, the decline of each subgradient has previous subgradient to decline
The influence in direction, this method helps speed up vector and declines to gradient is correctly oriented, thus make its convergence rate faster, with
Lower formula is weight and the formula that threshold value is updated with momentum.
Wherein,
υdωj=β * υdωj-1+(1-β)*dωj-1；
υdbj=β * υdbj-1+ (1- β) * dbj-1
In above-mentioned formula, ω is weight, ωjWeight when for iteration j, ωj-1Weight when for -1 iteration of jth；
B is threshold value, bjThreshold value when for iteration j, bj-1Threshold value when for -1 iteration of jth；α is learning rate, also uses LR
(Learning Rate) is indicated；vdωjThe speed in right value update direction, v when for iteration jdbjThreshold value when for iteration j
The speed of more new direction, partial derivative of the loss function to ω and b when calculation is respectively iteration j；vdωj-1For jth -1
The speed in right value update direction, υ when secondary iterationdbj-1The speed of threshold value more new direction when for -1 iteration of jth, calculation difference
Partial derivative of the loss function to ω and b when for -1 iteration of jth；β is the hyper parameter of self-setting, is generally defaulted as 0.9, here
According to default value.
Fig. 5, Fig. 6 show the process for introducing in stochastic gradient descent method and finding global optimum when the training of momentum front and back.Fig. 5
Decline for the gradient of no momentum, Fig. 6 is the gradient decline training method that momentum is added, terraced in Fig. 6 it can be seen from Fig. 5, Fig. 6
Momentum is introduced in degree descent method, an acceleration is increased in the direction of gradient decline, globe optimum can be found faster.
The present embodiment carries out failure to refrigeration system using deep neural network (Deep Neural Network, DNN) and examines
It is disconnected, using small lot momentum stochastic gradient descent (Stochastic Gradient Descent Momentum, SGDM) training
Method, each step gradient calculating only randomly select one group of sample in training set, improve its convergence rate.
Deep neural network model formations 25 is according to the foundation such as the structure of neural network, activation primitive, loss function depth
Spend neural network model.
Deep neural network model formations 25 is according to the foundation such as the structure of neural network, activation primitive, loss function depth
Neural network model is spent, determines the topological structure of deep neural network, which includes the input layer of deep neural network
Number, weight and threshold value.
Train epochs configuration part 26 is used to determine the train epochs of deep neural network.
(150,200,250,300 ...) are gradually tested from 100, and to find preferable train epochs, (judgment criteria is most
Rate of correct diagnosis is higher after test set input afterwards), so that it is determined that the train epochs of neural network.
Fault diagnosis model building portion 27 is trained rear structure for application training group data in deep neural network model
Build fault diagnosis model.
Start to train neural network, be optimized using simulated annealing (with 0.01 or 0.001 rate to study
Rate decays).
Learning rate Optimization Dept. 28 optimizes learning rate using simulated annealing, and specific flow chart is as shown in Figure 7.
Fig. 7 is the flow chart that simulated annealing carries out global optimizing to the learning rate of deep neural network, be first arranged compared with
Big learning rate, because initial random weight is far from optimal value.In the training process, a constant is introduced to cause learning rate
Disturbance, formula are as follows:
LRj=0.1*LRj-1
Learning rate is gradually reduced, to allow fine-grained weight to update.Initial learning rate is since 0.1, then index again
Decline, 0.01,0.001....
In formula, j is current iteration number, and j-1 is a upper the number of iterations, and LR represents Learning Rate, learning rate, LRj
For the learning rate of current iteration number, LRj-1For the learning rate of a upper the number of iterations.So that error precision rapidly drops to one
Smaller value.This process is referred to as simulated annealing, because it is similar to the metallurgy annealing process that molten metal slowly cools down, mould
Quasi- annealing optimization DNN reduces the time of DNN neural metwork training, improves neural metwork training speed, solves trained DNN
Neural network cannot obtain the problem of globally optimal solution.
In order to reach global optimum, the present embodiment uses simulated annealing optimization deep neural network (Simulated simultaneously
Annealing-Deep Neural Network, SA-DNN) model, Lai Youhua deep neural network learning rate, α is under gradient
It drops and is referred to as learning rate or step-length in algorithm, it is meant that the distance that each step is walked can be controlled, by d to guarantee not walk
It is too fast, miss minimum point.Also to guarantee not walk simultaneously is too slow, causes slowly to walk less than minimum point, α control is one by SA
It is a to make DNN neural network first with faster speed training with the gradually smaller value of frequency of training, when model training is to close to the overall situation
With the training of lesser learning rate when optimal.
If meeting network termination condition (Rule of judgment is set train epochs), tested to come using test set
Verify the performance of proposed diagnostic model.If not satisfied, then continuing that regularized learning algorithm is blunt to be terminated to network training.
When meeting termination condition, trained fault diagnosis model is obtained.
Fault diagnosis result generating unit 29 carries out failure using test group data in trained fault diagnosis model and examines
It is disconnected, obtain fault diagnosis result.
Fault diagnosis is carried out to test group data in S2 in trained fault diagnosis model, obtains fault diagnosis result.
The output form of diagnostic result are as follows: XX failure is diagnosed as XX failure, can be indicated with the form of confusion matrix.
Diagnosis side control unit 31 is for controlling each portion of above-mentioned refrigeration equipment trouble-shooter 20.
Refrigeration equipment trouble-shooter 20 is any one in fixed terminal and mobile terminal.Fixed terminal includes platform
Formula computer, mobile terminal include smart phone, tablet computer.
In embodiment, refrigeration equipment trouble-shooter 20 is desktop computer.
In the present embodiment, refrigeration equipment fault diagnosis system 100 is answered using simulated annealing SA optimization DNN deep neural network
It is diagnosed for breakdown of refrigeration system, the results showed that
As shown in figure 8, being thermodynamic chart of the SA-DNN compared with DNN model the number of hidden nodes and rate of correct diagnosis.
Wherein, figure a), figure be b) single hidden layer DNN the number of hidden nodes with accuracy compared with, figure is a), to scheme b) abscissa be implicit
Node layer number, ordinate are rate of correct diagnosis, from figure a), figure b) in it can be seen that the number of hidden nodes be 64 when, rate of correct diagnosis
Highest.
C), d) figure is two hidden-layer number of nodes compared with accuracy, c), d) figure abscissa be first hidden layer number of nodes,
Ordinate is the number of nodes of second hidden layer, and the depth of color indicates that rate of correct diagnosis, specific depth numerical value show right in figure
At the colour atla of side, the region as lines indicate in figure is preferable result.
A), c) figure is mapped by DNN deep neural network model.
B), d) figure is mapped by SA-DNN deep neural network model.
As shown in figure 9, SA-DNN, DNN are compared with BP neural network diagnosis performance.Abscissa is network structure, main vertical seat
It is designated as rate of correct diagnosis, secondary ordinate is that diagnosis is time-consuming.Abscissa the first two is BP model, and intermediate two are DNN model, rear two
A is SA-DNN model.BP1_18 indicates 18 node BP neural networks of single hidden layer, and BP2_25 indicates every layer of two hidden-layer 25 sections
Point BP neural network；DNN1_64 indicates 64 node DNN neural networks of single hidden layer, and DNN2_32 indicates every layer of two hidden-layer 32 sections
Point DNN neural network；SA-DNN1_128 indicates that simulated annealing optimization list hidden layer 128 nodes DNN, SA-DNN2_64 indicate mould
The network of quasi- annealing optimization 64 node of every layer of two hidden-layer.
As can be seen from the figure BP1_18 Model Diagnosis accuracy is minimum, and BP2_25 Model Diagnosis time-consuming is for up to 14.16,
Using time-consuming least model DNN1_64 as unit 1, the time-consuming of other models is the time-consuming multiple of model DNN1_64 diagnosis.
14.16 be 14.16 times that BP2_25 Model Diagnosis time-consuming is DNN1_64 diagnosis time-consuming.
The rate of correct diagnosis of SA-DNN2_64 is 99.3%, and rate of correct diagnosis is highest, diagnosis consumption in six models
When be 1.04.
To sum up the result shows that: SA-DNN result stablize, be not easy to fall into local minimum, more effectively realize refrigeration system
Fault diagnosis, and have excellent performance.
As shown in figure 3, the process of Simulated Anneal Algorithm Optimize deep neural network are as follows:
A1, data collection are simultaneously handled；
The number of nodes and the number of plies of deep neural network is arranged in A2；
A3 establishes deep neural network model according to the structure of neural network, activation primitive, loss function etc.；
A4 determines the train epochs of deep neural network；
A5, training deep neural network；
A6 optimizes learning rate using simulated annealing；
A7 judges to meet the condition terminated, if being judged as YES, into next step, if being judged as NO, returns to A6；
A8 obtains trained fault diagnosis model；
A9 carries out fault diagnosis using test group data in trained fault diagnosis model, obtains fault diagnosis knot
Fruit；
A10 terminates.
As shown in fig. 7, Simulated Anneal Algorithm Optimize learning rate process are as follows:
B1 generates initial learning rate LR；
B2 calculates loss function C；
B3, disturbance generate new learning rate LR_new；
B4 calculates Δ C=C (LR_new)-C (LR)；
B5 judges to meet the condition terminated, if being judged as YES, into next step, if being judged as NO, into B7；
B6 receives new explanation LR_new；
B7 receives new explanation LR_new with certain probability, into B8；
B8 judges to meet the condition terminated, if being judged as YES, into next step, if being judged as NO, returns to B3；
B9 returns to variable optimal learning rate；
B10 terminates.
The action and effect of embodiment
Easily sink into local minimum when for common BP neural network optimizing, and network structure is not sufficiently stable, every time training
Obtained result differs the problems such as larger and net training time is too long, in order to improve deep neural network (Deep Neural
Network, DNN) to breakdown of refrigeration system diagnosis efficiency and accuracy, the refrigeration equipment trouble-shooter of the present embodiment and
System application small lot momentum stochastic gradient descent (Stochastic Gradient Descent Momentum, SGDM) training
Method is trained deep neural network, and each step gradient calculating only randomly selects one group of sample in training set, improves it
Convergence rate.
In addition, in order to reach global optimum, the refrigeration equipment trouble-shooter and system of the present embodiment use mould simultaneously
Quasi- annealing optimization deep neural network (Simulatedannealing-Deep Neural Network, SA-DNN) model.Knot
Fruit shows: to seven quasi-representative failures of refrigeration system, SA-DNN greatly improves rate of correct diagnosis, and optimal network structure is
2 nodes of hidden layer 64, rate of correct diagnosis 99.3%, time-consuming diagnosis is only 3min50s.(Back is propagated compared to conventional counter
Propagation, BP) neural network, the refrigeration equipment trouble-shooter and system of the present embodiment have SA-DNN result steady
It is fixed, it is not easy to fall into local minimum, more effectively realizes breakdown of refrigeration system diagnosis, and the beneficial effect of function admirable.
Above embodiment is preferred case of the invention, the protection scope being not intended to limit the invention.
Claims (9)
1. a kind of refrigeration equipment trouble-shooter characterized by comprising
Management with storage unit, node and number of plies configuration part, deep neural network training department, deep neural network model formations,
Train epochs configuration part, fault diagnosis model building portion, learning rate Optimization Dept. and fault diagnosis result generating unit,
Wherein, the management is at least stored with set evidence and test group data with storage unit,
The node and number of plies configuration part are used to be arranged the number of nodes and the number of plies of deep neural network,
The deep neural network training department is using SGDM (small lot momentum stochastic gradient descent method) to the depth nerve net
Network is trained,
The deep neural network model formations establishes depth according to the structure of neural network, activation primitive, loss function etc.
Neural network model,
The train epochs configuration part is used to determine the train epochs of deep neural network,
Fault diagnosis model building portion in the deep neural network model using the set according to being instructed
Fault diagnosis model is constructed after white silk,
The learning rate Optimization Dept. optimizes learning rate using simulated annealing,
The fault diagnosis result generating unit carries out fault diagnosis to the test group data, obtains fault diagnosis result.
2. refrigeration equipment trouble-shooter according to claim 1, it is characterised in that:
Wherein, the expression formula of the loss function C are as follows:
In formula, x indicates that sample, y indicate actual value, and a indicates output valve, and n indicates the sum of sample.
3. refrigeration equipment trouble-shooter according to claim 1, it is characterised in that:
Wherein, the output form of the diagnostic result can be indicated with the form of confusion matrix.
4. refrigeration equipment trouble-shooter according to claim 1, it is characterised in that:
Wherein, the expression formula of the activation primitive σ are as follows:
In formula, y=ω1×x1+ω2×x2+…ωm×xm+b
ωmFor the weight of m-th of hidden layer node of a certain layer, xmFor the input value of m-th of hidden layer node of a certain layer, b is a certain
The threshold value of m-th of hidden layer node of layer.
5. a kind of refrigeration equipment fault diagnosis system characterized by comprising
Data acquisition device；And
Refrigeration equipment trouble-shooter, the refrigeration equipment trouble-shooter are described in any one in claim 1-4
Refrigeration equipment trouble-shooter,
Wherein, the management in the refrigeration equipment trouble-shooter is stored with to acquire from the data with storage unit and be filled
Set the data of acquisition.
6. refrigeration equipment fault diagnosis system according to claim 5, it is characterised in that:
Wherein, the data acquisition device and the refrigeration equipment trouble-shooter communicate to connect.
7. refrigeration equipment fault diagnosis system according to claim 5, it is characterised in that:
Wherein, the refrigeration equipment trouble-shooter is any one in fixed terminal and mobile terminal.
8. refrigeration equipment fault diagnosis system according to claim 7, it is characterised in that:
Wherein, the fixed terminal includes desktop computer.
9. refrigeration equipment fault diagnosis system according to claim 7, it is characterised in that:
Wherein, the mobile terminal includes smart phone, tablet computer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910148909.4A CN109948778A (en) | 2019-02-28 | 2019-02-28 | A kind of refrigeration equipment trouble-shooter and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910148909.4A CN109948778A (en) | 2019-02-28 | 2019-02-28 | A kind of refrigeration equipment trouble-shooter and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109948778A true CN109948778A (en) | 2019-06-28 |
Family
ID=67008141
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910148909.4A Pending CN109948778A (en) | 2019-02-28 | 2019-02-28 | A kind of refrigeration equipment trouble-shooter and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109948778A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110285532A (en) * | 2019-07-04 | 2019-09-27 | 中国工商银行股份有限公司 | Method for controlling machine room air conditioner, apparatus and system based on artificial intelligence |
| CN110543893A (en) * | 2019-08-07 | 2019-12-06 | 河海大学 | Microwave attenuation precipitation particle type identification method based on BP neural network |
| CN110567112A (en) * | 2019-09-16 | 2019-12-13 | 重庆美的通用制冷设备有限公司 | control method of air conditioning equipment, control system of air conditioning equipment and air conditioning system |
| CN111981635A (en) * | 2020-07-21 | 2020-11-24 | 沈阳安新自动化控制有限公司 | Central air conditioner fault prediction and diagnosis method adopting double intelligent algorithms |
| CN114442477A (en) * | 2022-04-11 | 2022-05-06 | 北京信云筑科技有限责任公司 | Equipment health management system based on Internet of things |
| CN114638149A (en) * | 2022-01-20 | 2022-06-17 | 佳都科技集团股份有限公司 | Fault diagnosis method and device based on convolution |
| CN115127192A (en) * | 2022-05-20 | 2022-09-30 | 中南大学 | Semi-supervised water chilling unit fault diagnosis method and system based on graph neural network |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103245524A (en) * | 2013-05-24 | 2013-08-14 | 南京大学 | Acoustic fault diagnosis method based on neural network |
| CN108334059A (en) * | 2018-02-26 | 2018-07-27 | 上海理工大学 | Based on particle cluster algorithm Optimized BP Neural Network Symbolic 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 |
| CN108764601A (en) * | 2018-04-03 | 2018-11-06 | 哈尔滨工业大学 | A kind of monitoring structural health conditions abnormal data diagnostic method based on computer vision and depth learning technology |
| CN108873695A (en) * | 2018-06-01 | 2018-11-23 | 哈尔滨理工大学 | The full Connection Neural Network control system of multilayer |
-
2019
- 2019-02-28 CN CN201910148909.4A patent/CN109948778A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103245524A (en) * | 2013-05-24 | 2013-08-14 | 南京大学 | Acoustic fault diagnosis method based on neural network |
| CN108334059A (en) * | 2018-02-26 | 2018-07-27 | 上海理工大学 | Based on particle cluster algorithm Optimized BP Neural Network Symbolic 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 |
| CN108764601A (en) * | 2018-04-03 | 2018-11-06 | 哈尔滨工业大学 | A kind of monitoring structural health conditions abnormal data diagnostic method based on computer vision and depth learning technology |
| CN108873695A (en) * | 2018-06-01 | 2018-11-23 | 哈尔滨理工大学 | The full Connection Neural Network control system of multilayer |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110285532A (en) * | 2019-07-04 | 2019-09-27 | 中国工商银行股份有限公司 | Method for controlling machine room air conditioner, apparatus and system based on artificial intelligence |
| CN110285532B (en) * | 2019-07-04 | 2021-07-30 | 中国工商银行股份有限公司 | Machine room air conditioner control method, device and system based on artificial intelligence |
| CN110543893A (en) * | 2019-08-07 | 2019-12-06 | 河海大学 | Microwave attenuation precipitation particle type identification method based on BP neural network |
| CN110567112A (en) * | 2019-09-16 | 2019-12-13 | 重庆美的通用制冷设备有限公司 | control method of air conditioning equipment, control system of air conditioning equipment and air conditioning system |
| CN111981635A (en) * | 2020-07-21 | 2020-11-24 | 沈阳安新自动化控制有限公司 | Central air conditioner fault prediction and diagnosis method adopting double intelligent algorithms |
| CN111981635B (en) * | 2020-07-21 | 2021-12-31 | 沈阳安新自动化控制有限公司 | Central air conditioner fault prediction and diagnosis method adopting double intelligent algorithms |
| CN114638149A (en) * | 2022-01-20 | 2022-06-17 | 佳都科技集团股份有限公司 | Fault diagnosis method and device based on convolution |
| CN114638149B (en) * | 2022-01-20 | 2023-08-08 | 佳都科技集团股份有限公司 | Convolution-based fault diagnosis method and device |
| CN114442477A (en) * | 2022-04-11 | 2022-05-06 | 北京信云筑科技有限责任公司 | Equipment health management system based on Internet of things |
| CN114442477B (en) * | 2022-04-11 | 2022-06-07 | 北京信云筑科技有限责任公司 | Equipment health management system based on Internet of things |
| CN115127192A (en) * | 2022-05-20 | 2022-09-30 | 中南大学 | Semi-supervised water chilling unit fault diagnosis method and system based on graph neural network |
| CN115127192B (en) * | 2022-05-20 | 2024-01-23 | 中南大学 | Semi-supervised water chilling unit fault diagnosis method and system based on graph neural network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109948778A (en) | A kind of refrigeration equipment trouble-shooter and system | |
| CN109902741A (en) | A kind of breakdown of refrigeration system diagnostic method | |
| CN108334059A (en) | Based on particle cluster algorithm Optimized BP Neural Network Symbolic fault diagnosis method | |
| CN102567391B (en) | Method and device for building classification forecasting mixed model | |
| CN108376286A (en) | A kind of equipment fault diagnosis apparatus and system | |
| CN110298601A (en) | A kind of real time business air control system of rule-based engine | |
| CN108494810A (en) | Network security situation prediction method, apparatus and system towards attack | |
| CN106327062A (en) | Method for evaluating state of power distribution network equipment | |
| CN110929918A (en) | 10kV feeder line fault prediction method based on CNN and LightGBM | |
| CN106021377A (en) | Information processing method and device implemented by computer | |
| CN106408687B (en) | A kind of automobile engine fault early warning method based on machine learning method | |
| CN108799844B (en) | Fuzzy set-based water supply network pressure monitoring point site selection method | |
| CN105357691A (en) | LTE (Long Term Evolution) wireless network user sensitive monitoring method and system | |
| CN110298518A (en) | A kind of fire rescue route planning method | |
| CN108894282A (en) | City planting ductwork operational safety dynamic early-warning method | |
| CN104484548B (en) | A kind of improved sequential Fault Diagnosis Strategy optimization method | |
| CN106569030A (en) | Alarm threshold optimizing method and device in electric energy metering abnormity diagnosis | |
| CN104598968A (en) | Fault diagnosis method of transformer | |
| WO2021219080A1 (en) | Federated learning model-based view display method, apparatus and device, and medium | |
| CN109815904A (en) | Fire identification method based on convolutional neural network | |
| CN109800995A (en) | A kind of grid equipment fault recognition method and system | |
| Lockhart et al. | Scission: Performance-driven and context-aware cloud-edge distribution of deep neural networks | |
| CN109634820A (en) | A kind of fault early warning method, relevant device and the system of the collaboration of cloud mobile terminal | |
| WO2024056051A1 (en) | Non-intrusive flexible load aggregation characteristic identification and optimization method, apparatus, and device | |
| CN106202113A (en) | It is applied to information providing method and the device forming a team to move |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190628 |