CN113627496A - Method, device, electronic equipment and readable storage medium for predicting fault of turnout switch machine - Google Patents
Method, device, electronic equipment and readable storage medium for predicting fault of turnout switch machine Download PDFInfo
- Publication number
- CN113627496A CN113627496A CN202110849795.3A CN202110849795A CN113627496A CN 113627496 A CN113627496 A CN 113627496A CN 202110849795 A CN202110849795 A CN 202110849795A CN 113627496 A CN113627496 A CN 113627496A
- Authority
- CN
- China
- Prior art keywords
- fault
- value
- switch machine
- recall
- prediction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 64
- 239000011159 matrix material Substances 0.000 claims abstract description 75
- 230000009471 action Effects 0.000 claims description 18
- 238000013145 classification model Methods 0.000 claims description 15
- 238000004590 computer program Methods 0.000 claims description 12
- 238000006243 chemical reaction Methods 0.000 claims description 9
- 238000012795 verification Methods 0.000 claims description 5
- 230000007257 malfunction Effects 0.000 claims 6
- 238000012545 processing Methods 0.000 abstract description 14
- 238000010586 diagram Methods 0.000 description 15
- 238000012423 maintenance Methods 0.000 description 11
- 230000008439 repair process Effects 0.000 description 10
- 238000004891 communication Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 7
- 230000006870 function Effects 0.000 description 6
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 238000010276 construction Methods 0.000 description 4
- 238000002405 diagnostic procedure Methods 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000007635 classification algorithm Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 238000013024 troubleshooting Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 230000036541 health Effects 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000001105 regulatory effect Effects 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24317—Piecewise classification, i.e. whereby each classification requires several discriminant rules
-
- 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
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24133—Distances to prototypes
- G06F18/24137—Distances to cluster centroïds
- G06F18/2414—Smoothing the distance, e.g. radial basis function networks [RBFN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/20—Ensemble learning
Abstract
The embodiment of the application provides a turnout point switch fault prediction method and device, electronic equipment and a readable storage medium, and relates to the technical field of rail transit. According to the turnout switch machine fault prediction method, preliminary fault prediction is carried out on a feature matrix corresponding to the turnout switch machine through a pre-constructed recall model, under the condition that the feature matrix is to-be-predicted fault data, the to-be-predicted fault data is input into a pre-trained fault prediction model, and fault prediction is carried out by using the fault prediction model, so that fault prediction is carried out on the turnout switch machine in a multi-stage processing mode, a complex industrial environment can be met, and the accuracy of fault prediction is improved.
Description
Technical Field
The present application relates to the field of rail transit technologies, and in particular, to a method and an apparatus for predicting a switch point machine failure, an electronic device, and a readable storage medium.
Background
The operation and maintenance of the turnout switch machine are always important links of subway cooperative operation and maintenance, and whether the switch machine can timely and reliably respond to system commands and realize turnout conversion is related to the driving safety and the life and property safety of passengers. At present, the operation and maintenance methods of the domestic turnout switch machine are generally divided into three categories: analytical model-based diagnostic methods, knowledge-based diagnostic methods, and regulatory-based rights management. The three methods belong to the categories of 'planned repair' and 'fault repair'.
The term "planned maintenance" means that after the turnout switch machine is used for a certain period of time, the turnout switch machine is maintained or replaced once no matter how the state of the switch machine is or whether the turnout switch machine is in failure.
"troubleshooting" refers to building a model base or knowledge base to expand the sensitivity to faults based on the fault phenomenon. Currently, the most applied method is to judge the health state of the switch machine by monitoring whether the condition data of the switch machine exceeds a threshold value in a mode of manually setting the threshold value. In the 'fault repairing' product partially adopting the artificial intelligence technology, the working state (healthy or in a certain fault) of the current switch machine is obtained by matching the machine learning or deep learning classification algorithm with the known fault label based on the working condition data.
However, the traditional 'plan repair' and 'fault repair' are time-consuming and labor-consuming, and meanwhile, the existing 'prediction repair' solution only adopts one model for prediction, so that the complex industrial environment is difficult to face, and the accuracy of model prediction is difficult to guarantee.
Disclosure of Invention
The embodiment of the application provides a turnout switch machine fault prediction method, a turnout switch machine fault prediction device, electronic equipment and a readable storage medium, so as to improve the problems.
According to a first aspect of embodiments of the present application, there is provided a switch machine failure prediction method, the method comprising:
acquiring a characteristic matrix corresponding to current data of a turnout switch machine;
recalling the feature matrix by using a pre-constructed recall model, and determining whether the feature matrix is fault data to be predicted;
and under the condition that the characteristic matrix is determined to be fault data to be predicted, inputting the fault data to be predicted into a fault prediction model which is trained in advance, and performing fault prediction by using the fault prediction model to obtain a fault prediction result.
In an alternative embodiment, the recall model includes at least one recall submodel; the method comprises the following steps of utilizing a pre-constructed recall model to recall the feature matrix, and determining whether the feature matrix is fault data to be predicted or not, wherein the steps comprise:
respectively utilizing each recall submodel to predict fault values of the characteristic matrix to obtain at least one fault value;
comparing each fault value with a preset threshold value, taking the fault value larger than the preset threshold value as a target fault value, and determining the number of the target fault values;
and when the number of the target fault values is larger than or equal to the preset number, determining that the feature matrix is the fault data to be predicted.
In an optional embodiment, the method further comprises a step of setting the preset threshold according to a preset condition, the step comprising:
for each recall submodel, predicting a fault value by utilizing a pre-constructed verification set to obtain at least one fault value;
calculating an F1 value and a recall rate of a turnout switch machine based on all the fault values and original threshold values;
and updating the original threshold value according to the F1 value and the recall rate to obtain a preset threshold value.
In an optional embodiment, the step of updating the original threshold according to the F1 value and the recall rate to obtain a preset threshold includes:
increasing the original threshold according to a preset fixed step length to obtain a new original threshold;
calculating a new FI value and a new recall based on the new original threshold;
the original threshold is increased according to a preset fixed step length to obtain a new original threshold, and the steps of calculating a new FI value and a new recall rate based on the new original threshold are executed again until the new original threshold is equal to 1 to obtain a plurality of new original thresholds, and a new F1 value and a new recall rate corresponding to each new original threshold;
acquiring a target F1 value and a new original threshold corresponding to a target recall ratio, and taking the new original threshold as a preset threshold, wherein the target F1 value and the target recall ratio are maximum values of a new F1 value and a new recall ratio corresponding to all the new original thresholds.
In an alternative embodiment, the fault prediction model comprises a classification model;
under the condition that the characteristic matrix is determined to be fault data to be predicted, inputting the fault data to be predicted into a fault prediction model trained in advance, and performing fault prediction by using the fault prediction model to obtain a fault prediction result, wherein the step of obtaining the fault prediction result comprises the following steps:
and inputting the fault data to be predicted into the classification model, and performing fault prediction by using the classification model to obtain a fault prediction result, wherein the fault prediction result represents whether the turnout switch machine fails within a future preset time.
In an alternative embodiment, the fault prediction model further comprises a time-to-failure prediction model;
under the condition that the fault prediction result represents that the turnout switch machine fails within the future preset time, inputting the data of the fault to be predicted corresponding to the turnout switch machine into the fault time prediction model, and predicting the fault time by using the fault time prediction model to obtain a fault time prediction result;
the failure time prediction result represents the time of failure of the turnout switch machine.
In an optional embodiment, the step of obtaining a feature matrix corresponding to current data of the switch machine includes:
acquiring current data of the turnout switch machine, wherein the turnout switch machine is an alternating current turnout switch machine, and the current data comprises current data corresponding to three-phase current;
dividing current data corresponding to each phase of current according to the action working condition of the turnout switch machine to obtain at least two pieces of sub-current data, wherein the action working condition comprises unlocking, conversion, locking and slow release;
extracting the maximum value, the average value, the standard value and the minimum value included by the sub-current data to obtain a first characteristic element;
acquiring the conversion time of the turnout switch machine corresponding to each sub-current data to obtain a second characteristic element;
and constructing a feature matrix based on the first feature element and the second feature element.
According to a second aspect of embodiments of the present application, there is provided a switch machine failure prediction device, the device including:
the acquisition module is used for acquiring a characteristic matrix corresponding to current data of the turnout switch machine;
the recall module is used for recalling the feature matrix by utilizing a pre-constructed recall model and determining whether the feature matrix is to-be-predicted fault data;
and the fault prediction module is used for inputting the fault data to be predicted into a fault prediction model which is trained in advance under the condition that the characteristic matrix is determined to be the fault data to be predicted, and performing fault prediction by using the fault prediction model to obtain a fault prediction result.
In an alternative embodiment, the recall model includes at least one recall submodel;
the recall module is used for predicting fault values of the feature matrix by using each recall submodel respectively to obtain at least one fault value; comparing each fault value with a preset threshold value, taking the fault value larger than the preset threshold value as a target fault value, and determining the number of the target fault values; and under the condition that the number of the target fault values is greater than or equal to the preset number, determining that the feature matrix is to-be-predicted fault data.
According to a third aspect of the embodiments of the present application, there is provided an electronic device, the electronic device includes a processor, a memory and a bus, the memory stores machine-readable instructions executable by the processor, when the electronic device is running, the processor and the memory communicate with each other through the bus, and the processor executes the machine-readable instructions to perform the steps of the switch machine failure prediction method described above.
According to a fourth aspect of the embodiments of the present application, there is provided a readable storage medium storing a computer program, which when executed, implements the steps of the method for predicting a failure of a point switch machine as described above.
The embodiment of the application provides a turnout switch machine fault prediction method, a turnout switch machine fault prediction device, electronic equipment and a readable storage medium.
In order to make the aforementioned objects, features and advantages of the present application more comprehensible, several embodiments accompanied with figures are described in detail below.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the application and together with the description serve to explain the application and not to limit the application. In the drawings:
fig. 1 is a block diagram of an electronic device according to an embodiment of the present disclosure.
Fig. 2 is a schematic flow chart of a method for predicting a fault of a point switch machine according to an embodiment of the present disclosure.
Fig. 3 is a flow chart illustrating sub-steps of a method for predicting a switch machine failure according to an embodiment of the present disclosure.
Fig. 4 is a flow chart illustrating the second sub-steps of a method for predicting a switch machine failure according to an embodiment of the present invention.
Fig. 5 is a third flow chart illustrating the sub-steps of a method for predicting a switch machine failure according to an embodiment of the present invention.
Fig. 6 is a network structure diagram of a failure time prediction model according to an embodiment of the present application.
Fig. 7 is a functional block diagram of a turnout switch machine failure prediction device according to an embodiment of the present application.
Icon: 100-an electronic device; 110-a memory; 120-a processor; 130-turnout switch machine failure prediction device; 131-an acquisition module; 132-a recall module; 133-a failure prediction module; 140-a communication unit.
Detailed Description
As introduced in the background art, the operation and maintenance of the turnout switch machine is always an important link of the coordinated operation and maintenance of the subway, and whether the switch machine can timely and reliably respond to system commands and realize turnout switching is related to the driving safety and the life and property safety of passengers. At present, the operation and maintenance methods of the domestic turnout switch machine are generally divided into three categories: analytical model-based diagnostic methods, knowledge-based diagnostic methods, and regulatory-based rights management. The three methods belong to the categories of 'planned repair' and 'fault repair'.
The term "planned maintenance" means that after the turnout switch machine is used for a certain period of time, the turnout switch machine is maintained or replaced once no matter how the state of the switch machine is or whether the turnout switch machine is in failure.
"troubleshooting" refers to building a model base or knowledge base to expand the sensitivity to faults based on the fault phenomenon. Currently, the most applied method is to judge the health state of the switch machine by monitoring whether the condition data of the switch machine exceeds a threshold value in a mode of manually setting the threshold value. In the 'fault repairing' product partially adopting the artificial intelligence technology, the working state (healthy or in a certain fault) of the current switch machine is obtained by matching the machine learning or deep learning classification algorithm with the known fault label based on the working condition data.
In other words, in the process of implementing the application, the inventor finds that the traditional 'plan repair' and 'fault repair' are time-consuming and labor-consuming, and meanwhile, the existing 'prediction repair' solution only adopts one model for prediction, so that the complex industrial environment is difficult to be faced, and the accuracy of model prediction is difficult to be ensured.
In view of the above problems, embodiments of the present application provide a method, an apparatus, an electronic device, and a readable storage medium for predicting a fault of a turnout switch machine, where the method obtains a feature matrix corresponding to current data of the turnout switch machine. And recalling the feature matrix by using a pre-constructed recall model to determine whether the feature matrix is the fault data to be predicted. And under the condition that the characteristic matrix is determined to be the fault data to be predicted, inputting the fault data to be predicted into a fault prediction model which is trained in advance, and performing fault prediction by using the fault prediction model to obtain a fault prediction result. Therefore, the fault prediction is carried out on the turnout switch machine in a multi-stage processing mode, the complex industrial environment can be met, and the accuracy of the fault prediction is improved.
The scheme in the embodiment of the present application may be implemented by using various computer languages, for example, object-oriented programming languages Java, C + +, and JavaScript.
The above prior art solutions have drawbacks that are the results of practical and careful study, and therefore, the discovery process of the above problems and the solutions proposed by the following embodiments of the present application to the above problems should be the contributions of the applicant to the present application in the course of the present application.
In order to make the technical solutions and advantages of the embodiments of the present application more apparent, the following further detailed description of the exemplary embodiments of the present application with reference to the accompanying drawings makes it clear that the described embodiments are only a part of the embodiments of the present application, and are not exhaustive of all embodiments. It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict.
Referring to fig. 1, fig. 1 is a block diagram of an electronic device 100 according to an embodiment of the present disclosure. The apparatus may include a processor 120, a memory 110, a point switch failure prediction device 130, and a communication unit 140, where the memory 110 stores machine readable instructions executable by the processor 120, and when the electronic apparatus 100 operates, the processor 120 and the memory 110 communicate with each other through a bus, and the processor 120 executes the machine readable instructions and performs a point switch failure prediction method.
The elements of the memory 110, the processor 120 and the communication unit 140 are electrically connected to each other directly or indirectly to realize the transmission or interaction of signals.
For example, the components may be electrically connected to each other via one or more communication buses or signal lines. The switch machine failure prediction device 130 includes at least one software function module that can be stored in the memory 110 in the form of software or firmware (firmware). The processor 120 is configured to execute executable modules stored in the memory 110, such as software functional modules or computer programs included in the switch machine failure prediction device 130.
The Memory 110 may be, but is not limited to, a Random Access Memory (RAM), a Read Only Memory (ROM), a Programmable Read-Only Memory (PROM), an Erasable Read-Only Memory (EPROM), an electrically Erasable Read-Only Memory (EEPROM), and the like.
The processor 120 may be an integrated circuit chip having signal processing capabilities. The Processor 120 may be a general-purpose Processor, and includes a Central Processing Unit (CPU), a Network Processor (NP), and so on.
But may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware components. The various methods, steps, and logic blocks disclosed in the embodiments of the present application may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
In the embodiment of the present application, the memory 110 is used for storing a program, and the processor 120 is used for executing the program after receiving the execution instruction. The method defined by the process disclosed in any of the embodiments of the present application can be applied to the processor 120, or implemented by the processor 120.
The communication unit 140 is used to establish a communication connection between the electronic apparatus 100 and another electronic apparatus via a network, and to transmit and receive data via the network.
In some embodiments, the network may be any type of wired or wireless network, or combination thereof. Merely by way of example, the Network may include a wired Network, a Wireless Network, a fiber optic Network, a telecommunications Network, an intranet, the internet, a Local Area Network (LAN), a Wide Area Network (WAN), a Wireless Local Area Network (WLAN), a Metropolitan Area Network (MAN), a Wide Area Network (WAN), a Public Switched Telephone Network (PSTN), a bluetooth Network, a ZigBee Network, a Near Field Communication (NFC) Network, or the like, or any combination thereof.
In the embodiment of the present application, the electronic device 100 may be, but is not limited to, a smart phone, a personal computer, a tablet computer, or the like having a processing function.
It will be appreciated that the configuration shown in figure 1 is merely illustrative. Electronic device 100 may also have more or fewer components than shown in FIG. 1, or a different configuration than shown in FIG. 1. The components shown in fig. 1 may be implemented in hardware, software, or a combination thereof.
The steps of the method for predicting the fault of the point switch provided by the embodiment of the present application are described in detail below based on the block diagram of the electronic device 100 shown in fig. 1.
Referring to fig. 2, fig. 2 is a schematic flow chart of a method for predicting a fault of a point switch machine according to an embodiment of the present disclosure.
And step S1, acquiring a feature matrix corresponding to the current data of the turnout switch machine.
The turnout switch machine is important signal basic equipment used for reliably converting the position of a turnout, changing the opening direction of the turnout, locking a turnout switch blade and reflecting the position of the turnout. The switch machine can adopt an alternating current switch machine and can also adopt a direct current switch machine. The current data may be current data of the switch machine.
And step S2, recalling the feature matrix by using a pre-constructed recall model, and determining whether the feature matrix is to-be-predicted fault data.
The recall processing can be understood as preliminary fault pre-judgment in the actual application stage, and whether the characteristic matrix is fault data to be predicted is determined. In the model building stage, the recall processing may be understood as recalling a positive sample (i.e., a fault sample) from a plurality of training samples to filter out samples without judgment value, so that the positive and negative samples are balanced.
And step S3, inputting the fault data to be predicted into a fault prediction model trained in advance under the condition that the characteristic matrix is determined to be the fault data to be predicted, and performing fault prediction by using the fault prediction model to obtain a fault prediction result.
According to the turnout switch machine fault prediction method provided by the embodiment of the application, the preliminary fault pre-judgment is carried out on the feature matrix through the pre-constructed recall model, after the feature matrix is determined to be the fault data to be predicted, the fault data to be predicted is input into the pre-trained fault prediction model, and the fault prediction is carried out by using the fault prediction model.
Because the acquired current data, namely the current data of the turnout switch machine, are continuous data, in order to reduce the calculation amount and ensure that the data to be predicted are representative so as to improve the accuracy of prediction, the characteristic construction needs to be carried out on the current data in the embodiment of the application, and the construction mode can be realized through the following modes:
in an alternative embodiment, please refer to fig. 3 in combination, fig. 3 is a flow chart illustrating the sub-steps of a method for predicting a switch machine fault according to an embodiment of the present disclosure. In step S1 shown in fig. 2, obtaining the feature matrix corresponding to the current data of the switch machine can be implemented by the following steps S11-S15:
and step S11, acquiring current data of a turnout switch machine, wherein the turnout switch machine is an alternating current turnout switch machine, and the current data comprises current data corresponding to three-phase current.
And step S12, dividing the current data corresponding to each phase of current according to the action working conditions of the turnout switch machine to obtain at least two parts of sub-current data, wherein the action working conditions comprise unlocking, conversion, locking and slow release.
Step S13, for each sub-current data, extracting a maximum value, an average value, a standard value, and a minimum value included in the sub-current data to obtain a first feature element.
Step S14, obtaining the conversion time of the turnout switch machine corresponding to each sub-current data to obtain a second characteristic element;
step S15, a feature matrix is constructed based on the first feature elements and the second feature elements.
Taking a turnout switch machine as an alternating-current turnout switch machine as an example, the process of establishing characteristics of the current data and obtaining the characteristic matrix is elaborated in detail.
According to the action working condition of the turnout switch machine, the turnout switch machine can be divided into four action stages of unlocking, converting, locking and releasing. And dividing the current data according to 4 time sections corresponding to the action working conditions of the turnout switch machine to obtain sub-current data corresponding to the four action stages respectively.
Meanwhile, a power supply used by the alternating current turnout switch machine is three-phase power, and current data of the alternating current turnout switch machine also comprises three phases, so that one-step characteristic construction needs to be carried out on sub-current data included by each phase of current data. That is, the maximum value, the average value, the standard value, and the minimum value included in the sub-current data are extracted to obtain the first feature element. At this time, the number of the characteristic elements is 48 in total, namely 3 × 4, wherein "3" indicates A, B, C three alternating current phases in total, the first "4" indicates four action sections, and the second "4" indicates the maximum value, the average value, the standard value and the minimum value included in each action phase.
And acquiring the conversion time of the turnout switch machine corresponding to each sub-current data to obtain a second characteristic element, wherein the conversion time of the turnout switch machine is the time of the turnout switch machine for action, such as the time of the turnout for executing opening action or the time of the turnout for executing closing action. For example, the switch machine is installed on the place A, the default road direction is the first direction, the first direction can lead to the city B, and after the switch machine is switched by executing the opening action, the road direction is changed into the second direction, and the second direction can lead to the city C. The switch machine performs the opening operation at 19 o' clock 30/7/1/2021, and the second characteristic element may be 2021711930.
And finally, constructing a feature matrix based on the first feature elements and the second feature elements. It should be noted that, in the embodiment of the present application, the order of the feature elements in the constructed feature matrix is not required, and only needs to be the same as the order of the feature elements in the feature matrix constructed in the training stage.
Therefore, according to the action process of the turnout switch machine, the current data of each phase is divided into four stages, the maximum value, the average value, the standard value and the minimum value of each stage are extracted, and finally the feature matrix comprising 49 elements is obtained, so that the fault features with weak differences relative to the whole sample in each action stage are captured, the calculated amount is reduced, and the precision and the speed of fault prediction are improved.
In an alternative embodiment, please refer to fig. 4 in combination, fig. 4 is a second flowchart illustrating the sub-steps of a method for predicting a switch machine fault according to an embodiment of the present disclosure. Step S2 shown in fig. 2, recalling the feature matrix by using a pre-constructed recall model, and determining whether the feature matrix is the fault data to be predicted, can be realized by the following steps S21-S23:
and step S21, respectively utilizing each recall submodel to predict fault values of the feature matrix to obtain at least one fault value.
Step S22, comparing each fault value with a preset threshold value, regarding the fault value greater than the preset threshold value as a target fault value, and determining the number of the target fault values.
In step S23, when the number of the target failure values is greater than or equal to the preset number, it is determined that the feature matrix is the failure data to be predicted.
The recall submodel may be implemented by an XGBoost algorithm. XGBoost is an ensemble learning algorithm composed of k base models, as shown in the following equation,
wherein f istFor recalling a submodel (i.e., a base model), k are included.Is the predicted value (i.e., fault value) of the ith sample (i.e., feature matrix).
Wherein the recall model comprises at least one recall submodel. That is, the number k may be 1, 2, 3, 4, 5 …
Meanwhile, the preset threshold is set in advance, the preset number is generally half of the number of the recall submodels, for example, if the number of the recall submodels is 1, the preset number is taken as 1. If the number of the recall submodels is 5, the preset number is 3. If the number of the recall submodels is 10, the preset number may be 6, 7, 8 or 10.
Taking the number of the recall submodels as 1, the preset threshold value as 0.5 and the preset number as 1 as examples, the recall submodels are used for predicting the fault value of the feature matrix, and the fault value is 0.6. The fault value 0.6 is compared with a preset threshold value 0.5, and since the fault value is greater than the preset threshold value, the fault value can be determined as a target fault value, and the number of the fault values is 1. And determining the characteristic matrix as the fault data to be predicted as the number of the target fault values is equal to the preset number.
Taking the number of the recall submodels as 3, the preset threshold value as 0.6 and the preset number as 2 as examples, respectively utilizing each recall submodel to carry out fault value prediction on the feature matrix, and obtaining the fault values as 0.5, 0.65 and 0.7 respectively. Comparing the fault values 0.5, 0.65 and 0.7 with the preset threshold value 0.6 respectively, the target fault values greater than the preset threshold value are 0.65 and 0.7, and the number of the target fault values is 2. And determining the characteristic matrix as the fault data to be predicted as the number of the target fault values is equal to the preset number.
Therefore, in practical application, the embodiment of the application firstly predicts the fault value of the feature matrix through the pre-constructed recall model so as to perform fault troubleshooting in advance, and determines the feature matrix as the fault data to be predicted on the premise that the quantity of the fault value larger than the preset threshold is larger than a certain numerical value, so that the prediction accuracy is improved.
Further, the method for constructing the recall model may refer to the prior art, and is not described herein in detail, but in order to improve the accuracy of the fault prediction, a method for updating the original threshold according to the F1 value and the recall rate is further adopted in the stage of constructing the recall model to obtain the preset threshold, which is described in detail below.
In an optional embodiment, the method further comprises the step of setting a preset threshold according to a preset condition, the step comprising:
and for each recall sub-model, predicting a fault value by utilizing a pre-constructed verification set to obtain at least one fault value. Based on all fault values and original thresholds, F1 values and switch machine recall rates are calculated. And updating the original threshold value according to the F1 value and the recall rate to obtain a preset threshold value.
The verification set is constructed by feature matrixes corresponding to historical current data of a plurality of turnout switch machines.
And after the recall submodel is constructed, utilizing the recall submodel to predict the fault values of the verification set to obtain at least one fault value. Its original threshold is a manually preset value, typically 0.5. The failure values are compared with the original threshold values one by one to calculate the F1 value and the recall rate of the points switches.
The recall ratio is calculated as follows:
wherein Recall is the Recall rate, TP is the number of correct predictions of the Recall submodel, and FN is the number of incorrect predictions of the Recall submodel.
The F1 value was calculated as follows:
f1 value ═ accuracy · recall · 2/(accuracy + recall)
Further, the original threshold may be updated according to the F1 value and the recall rate by increasing or decreasing the original threshold according to a fixed step size, and taking the value corresponding to the highest F1 value and the highest recall rate as the preset threshold.
And increasing the original threshold according to a preset fixed step length to obtain a new original threshold.
Based on the new original threshold, a new FI value and a new recall are calculated.
And increasing the original threshold according to the preset fixed step length to obtain a new original threshold, and calculating a new FI value and a new recall rate based on the new original threshold until the new original threshold is equal to 1 to obtain a plurality of new original thresholds, and a new F1 value and a new recall rate corresponding to each new original threshold.
And acquiring a target F1 value and a new original threshold corresponding to the target recall rate, and taking the new original threshold as a preset threshold, wherein the target F1 value and the target recall rate are the maximum values of all new original thresholds corresponding to the new F1 value and the new recall rate.
For example, the fixed step size may be 0.1 and the original threshold may be 0.5. Assuming that the corresponding calculated F1 value and recall rate are the highest after the original threshold is increased twice according to the fixed step length (i.e., the updated original threshold is 0.7), 0.7 is used as the preset threshold.
Therefore, the method and the device have the advantages that the proper preset threshold is selected by combining the F1 value and the recall rate, so that the accuracy of the recall model for determining whether the feature matrix is the fault data to be predicted is improved.
As an alternative implementation, please refer to fig. 5 in combination, fig. 5 is a third flow chart of the sub-steps of the method for predicting a switch machine fault according to the embodiment of the present application. In step S3 shown in fig. 2, when it is determined that the feature matrix is to-be-predicted fault data, the to-be-predicted fault data is input into a fault prediction model trained in advance, and fault prediction is performed by using the fault prediction model, so as to obtain a fault prediction result, which can be implemented as follows:
and step S31, inputting the fault data to be predicted into the classification model, and performing fault prediction by using the classification model to obtain a fault prediction result, wherein the fault prediction result represents whether the turnout switch machine fails within a future preset time.
I.e. the fault prediction model comprises a classification model. The classification model can also be constructed by using the XGBoost algorithm, and of course, the classification model can also be implemented by using other common classification algorithms, and the construction process thereof can refer to the prior art and is not described herein again. The predetermined time in the future may be 1 day, 2 days, 3 days, 4 days, 5 days, or the like.
Therefore, after the characteristic matrix is subjected to fault pre-judgment processing through the recall model, fault prediction is carried out through the classification model again, and the accuracy of fault prediction of the turnout switch machine is improved. The maintenance schedule can be customized by maintenance personnel in time, and the economic loss and the safety risk caused by the failure of the turnout switch machine are reduced.
On the basis, in order to further improve the accuracy of fault prediction, on the basis that the fault prediction model comprises a classification model and the classification model is used for fault prediction, the embodiment of the application further adopts a fault time prediction model, and under the condition that the fault prediction result given by the classification model represents that the turnout switch opportunity has a fault, the fault time prediction model is used for predicting the specific time of the turnout switch machine with the fault, which is elaborated in detail below:
and step S32, under the condition that the fault prediction result represents that the turnout switch machine has a fault within the future preset time, inputting the data of the fault to be predicted corresponding to the turnout switch machine into a fault time prediction model, and predicting the fault time by using the fault time prediction model to obtain a fault time prediction result which represents the time when the turnout switch machine has the fault.
Referring to fig. 6, fig. 6 is a network structure diagram of a failure time prediction model according to an embodiment of the present disclosure. The fault time prediction model can be a neural network with residual connection, and comprises an input layer (Inputs) independent full-connection layer (Dense layer), two intermediate layers and an output layer, wherein each intermediate layer comprises a batch normalization processing layer (Batchnormalization), an under-drawn fit layer (Droput) and a full-connection layer (Dense layer), and the independent full-connection layer is connected with the full-connection layer of the second intermediate layer to form a residual structure.
As a possible case, as shown in fig. 6, the prediction result of the failure time output by the output layer is "00111", which indicates that the switch machine corresponding to the failure data to be predicted will fail on the third day.
It should be noted that, the training method of the fault time prediction model may refer to the prior art, and is not described herein again.
So, this application embodiment can obtain the fault time of the switch machine that will break down through this fault time prediction model to in time carry out emergent record according to the prediction time for maintenance personal, reduce economic loss and the safety risk because of the switch machine breaks down and brings.
Based on the same inventive concept, please refer to fig. 7, and fig. 7 is a functional block diagram of a device for predicting a fault of a point switch machine according to an embodiment of the present invention. In the embodiment of the present application, there is provided a turnout switch machine failure prediction device 130 corresponding to the turnout switch machine failure prediction method shown in fig. 2, the device including:
the obtaining module 131 is configured to obtain a feature matrix corresponding to current data of the point switch machine.
The recall module 132 is configured to perform recall processing on the feature matrix by using a pre-constructed recall model, and determine whether the feature matrix is to-be-predicted fault data.
And the fault prediction module 133 is configured to, when it is determined that the feature matrix is to-be-predicted fault data, input the to-be-predicted fault data into a pre-trained fault prediction model, and perform fault prediction by using the fault prediction model to obtain a fault prediction result.
The recall model comprises at least one recall submodel;
the recall module 132 is configured to perform fault value prediction on the feature matrix by using each recall submodel, so as to obtain at least one fault value; comparing each fault value with a preset threshold value, taking the fault value larger than the preset threshold value as a target fault value, and determining the number of the target fault values; and under the condition that the number of the target fault values is greater than or equal to the preset number, determining that the feature matrix is to-be-predicted fault data.
Because the principle of solving the problem of the device in the embodiment of the application is similar to the method for predicting the fault of the turnout switch machine in the embodiment of the application, the implementation principle of the device can be referred to the implementation principle of the method, and repeated details are not repeated.
The present invention also provides a readable storage medium, in which a computer program is stored, and the computer program is executed to implement the steps of the method for predicting the fault of the point switch machine.
To sum up, according to the turnout switch machine fault prediction method, the turnout switch machine fault prediction device, the electronic device and the readable storage medium provided by the embodiment of the application, the preliminary fault pre-judgment is performed on the feature matrix through the pre-established recall model, after the feature matrix is determined to be the fault data to be predicted, the fault data to be predicted is input into the pre-trained fault prediction model, and the fault prediction is performed by using the fault prediction model.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While the preferred embodiments of the present application have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all alterations and modifications as fall within the scope of the application.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present application without departing from the spirit and scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is intended to include such modifications and variations as well.
Claims (11)
1. A method of predicting a malfunction of a switch machine, the method comprising:
acquiring a characteristic matrix corresponding to current data of a turnout switch machine;
recalling the feature matrix by using a pre-constructed recall model, and determining whether the feature matrix is fault data to be predicted;
and under the condition that the characteristic matrix is determined to be fault data to be predicted, inputting the fault data to be predicted into a fault prediction model which is trained in advance, and performing fault prediction by using the fault prediction model to obtain a fault prediction result.
2. The point switch machine failure prediction method of claim 1, wherein the recall model comprises at least one recall submodel; the method comprises the following steps of utilizing a pre-constructed recall model to recall the feature matrix, and determining whether the feature matrix is fault data to be predicted or not, wherein the steps comprise:
respectively utilizing each recall submodel to predict fault values of the characteristic matrix to obtain at least one fault value;
comparing each fault value with a preset threshold value, taking the fault value larger than the preset threshold value as a target fault value, and determining the number of the target fault values;
and under the condition that the number of the target fault values is greater than or equal to the preset number, determining that the feature matrix is to-be-predicted fault data.
3. The method of predicting a malfunction of a point switch machine as claimed in claim 2, further comprising the step of setting the preset threshold value according to a preset condition, the step comprising:
for each recall submodel, predicting a fault value by utilizing a pre-constructed verification set to obtain at least one fault value;
calculating an F1 value and a recall rate of a turnout switch machine based on all the fault values and original threshold values;
and updating the original threshold value according to the F1 value and the recall rate to obtain a preset threshold value.
4. The method of predicting the malfunction of a switch machine as claimed in claim 3, wherein the step of updating the original threshold according to the F1 value and the recall ratio to obtain a preset threshold comprises:
increasing the original threshold according to a preset fixed step length to obtain a new original threshold;
calculating a new FI value and a new recall based on the new original threshold;
the original threshold is increased according to a preset fixed step length to obtain a new original threshold, and the steps of calculating a new FI value and a new recall rate based on the new original threshold are executed again until the new original threshold is equal to 1 to obtain a plurality of new original thresholds, and a new F1 value and a new recall rate corresponding to each new original threshold;
acquiring a target F1 value and a new original threshold corresponding to a target recall ratio, and taking the new original threshold as a preset threshold, wherein the target F1 value and the target recall ratio are maximum values of a new F1 value and a new recall ratio corresponding to all the new original thresholds.
5. The method of predicting a malfunction of a switch machine as claimed in claim 1, wherein said malfunction prediction model comprises a classification model;
inputting the fault data to be predicted into a pre-trained fault prediction model, and performing fault prediction by using the fault prediction model to obtain a fault prediction result, wherein the step of obtaining the fault prediction result comprises the following steps:
and inputting the fault data to be predicted into the classification model, and performing fault prediction by using the classification model to obtain a fault prediction result, wherein the fault prediction result represents whether the turnout switch machine fails within a future preset time.
6. The point switch machine failure prediction method of claim 5, wherein the failure prediction model further comprises a time-to-failure prediction model;
under the condition that the fault prediction result represents that the turnout switch machine fails within the future preset time, inputting the data of the fault to be predicted corresponding to the turnout switch machine into the fault time prediction model, and predicting the fault time by using the fault time prediction model to obtain a fault time prediction result;
the failure time prediction result represents the time of failure of the turnout switch machine.
7. The method of predicting a malfunction of a point switch machine according to claim 1, wherein the step of obtaining a feature matrix corresponding to current data of the point switch machine includes:
acquiring current data of the turnout switch machine, wherein the turnout switch machine is an alternating current turnout switch machine, and the current data comprises current data corresponding to three-phase current;
dividing current data corresponding to each phase of current according to the action working condition of the turnout switch machine to obtain at least two pieces of sub-current data, wherein the action working condition comprises unlocking, conversion, locking and slow release;
extracting the maximum value, the average value, the standard value and the minimum value included by the sub-current data to obtain a first characteristic element;
acquiring the conversion time of the turnout switch machine corresponding to each sub-current data to obtain a second characteristic element;
and constructing a feature matrix based on the first feature element and the second feature element.
8. A switch machine failure prediction device, characterized in that the device comprises:
the acquisition module is used for acquiring a characteristic matrix corresponding to current data of the turnout switch machine;
the recall module is used for recalling the feature matrix by utilizing a pre-constructed recall model and determining whether the feature matrix is to-be-predicted fault data;
and the fault prediction module is used for inputting the fault data to be predicted into a fault prediction model which is trained in advance under the condition that the characteristic matrix is determined to be the fault data to be predicted, and performing fault prediction by using the fault prediction model to obtain a fault prediction result.
9. The switch machine failure prediction device of claim 8, wherein the recall model comprises at least one recall submodel;
the recall module is used for predicting fault values of the feature matrix by using each recall submodel respectively to obtain at least one fault value; comparing each fault value with a preset threshold value, taking the fault value larger than the preset threshold value as a target fault value, and determining the number of the target fault values; and under the condition that the number of the target fault values is greater than or equal to the preset number, determining that the feature matrix is to-be-predicted fault data.
10. An electronic device, comprising a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, when the electronic device is running, the processor and the memory communicate with each other via the bus, and the processor executes the machine-readable instructions to perform the steps of the point switch failure prediction method according to any one of claims 1-7.
11. A readable storage medium storing a computer program which, when executed, implements the steps of the method for predicting a failure of a switch machine as claimed in any one of claims 1 to 7.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110849795.3A CN113627496A (en) | 2021-07-27 | 2021-07-27 | Method, device, electronic equipment and readable storage medium for predicting fault of turnout switch machine |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110849795.3A CN113627496A (en) | 2021-07-27 | 2021-07-27 | Method, device, electronic equipment and readable storage medium for predicting fault of turnout switch machine |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113627496A true CN113627496A (en) | 2021-11-09 |
Family
ID=78381073
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110849795.3A Pending CN113627496A (en) | 2021-07-27 | 2021-07-27 | Method, device, electronic equipment and readable storage medium for predicting fault of turnout switch machine |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113627496A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114692969A (en) * | 2022-03-29 | 2022-07-01 | 西门子交通技术(北京)有限公司 | Point switch failure prediction method, device, electronic equipment and storage medium |
Citations (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2002236904A (en) * | 2001-02-08 | 2002-08-23 | Sony Corp | Data processor and processing method, recording medium and program |
CN101893667A (en) * | 2010-05-26 | 2010-11-24 | 广州市地下铁道总公司 | Fault detection system of AC turnout switch machine and method thereof |
CN103632211A (en) * | 2013-12-06 | 2014-03-12 | 清华大学 | Motor vehicle fault pre-warning and callback prediction system |
CN106976468A (en) * | 2017-03-09 | 2017-07-25 | 南京理工大学 | A kind of switch breakdown diagnostic method based on DWT and C SVM |
CN107215357A (en) * | 2017-05-25 | 2017-09-29 | 同济大学 | A kind of switch breakdown Forecasting Methodology |
CN107451004A (en) * | 2017-07-01 | 2017-12-08 | 南京理工大学 | A kind of switch breakdown diagnostic method based on qualitiative trends analysis |
CN110837718A (en) * | 2019-11-07 | 2020-02-25 | 交控科技股份有限公司 | Turnout fault detection method and device, electronic equipment and storage medium |
CN110866655A (en) * | 2019-11-25 | 2020-03-06 | 武汉地铁运营有限公司 | Intelligent turnout jamming fault early warning method based on power numerical analysis |
CN111275503A (en) * | 2020-03-20 | 2020-06-12 | 京东数字科技控股有限公司 | Data processing method and device for acquiring lost user recall success rate |
CN111432206A (en) * | 2020-04-24 | 2020-07-17 | 腾讯科技(北京)有限公司 | Video definition processing method and device based on artificial intelligence and electronic equipment |
CN111444428A (en) * | 2020-03-27 | 2020-07-24 | 腾讯科技(深圳)有限公司 | Information recommendation method and device based on artificial intelligence, electronic equipment and storage medium |
CN111538908A (en) * | 2020-06-22 | 2020-08-14 | 腾讯科技(深圳)有限公司 | Search ranking method and device, computer equipment and storage medium |
CN111860971A (en) * | 2020-06-29 | 2020-10-30 | 交控科技股份有限公司 | Method and device for predicting residual life of turnout switch machine |
CN112036505A (en) * | 2020-09-17 | 2020-12-04 | 广西交控智维科技发展有限公司 | Method and device for determining equipment state of turnout switch machine and electronic equipment |
CN112150206A (en) * | 2020-09-28 | 2020-12-29 | 京东数字科技控股股份有限公司 | Method and equipment for predicting user interested article |
CN112215629A (en) * | 2019-07-09 | 2021-01-12 | 百度在线网络技术(北京)有限公司 | Multi-target advertisement generation system and method based on construction countermeasure sample |
CN112348170A (en) * | 2020-11-10 | 2021-02-09 | 交控科技股份有限公司 | Fault diagnosis method and system for turnout switch machine |
CN112477917A (en) * | 2020-11-10 | 2021-03-12 | 交控科技股份有限公司 | Turnout fault detection method and device, electronic equipment and storage medium |
WO2021109578A1 (en) * | 2019-12-02 | 2021-06-10 | 北京天元创新科技有限公司 | Method and apparatus for alarm prediction during service operation and maintenance, and electronic device |
WO2021121106A1 (en) * | 2019-12-20 | 2021-06-24 | 深圳前海微众银行股份有限公司 | Federated learning-based personalized recommendation method, apparatus and device, and medium |
-
2021
- 2021-07-27 CN CN202110849795.3A patent/CN113627496A/en active Pending
Patent Citations (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2002236904A (en) * | 2001-02-08 | 2002-08-23 | Sony Corp | Data processor and processing method, recording medium and program |
CN101893667A (en) * | 2010-05-26 | 2010-11-24 | 广州市地下铁道总公司 | Fault detection system of AC turnout switch machine and method thereof |
CN103632211A (en) * | 2013-12-06 | 2014-03-12 | 清华大学 | Motor vehicle fault pre-warning and callback prediction system |
CN106976468A (en) * | 2017-03-09 | 2017-07-25 | 南京理工大学 | A kind of switch breakdown diagnostic method based on DWT and C SVM |
CN107215357A (en) * | 2017-05-25 | 2017-09-29 | 同济大学 | A kind of switch breakdown Forecasting Methodology |
CN107451004A (en) * | 2017-07-01 | 2017-12-08 | 南京理工大学 | A kind of switch breakdown diagnostic method based on qualitiative trends analysis |
CN112215629A (en) * | 2019-07-09 | 2021-01-12 | 百度在线网络技术(北京)有限公司 | Multi-target advertisement generation system and method based on construction countermeasure sample |
CN110837718A (en) * | 2019-11-07 | 2020-02-25 | 交控科技股份有限公司 | Turnout fault detection method and device, electronic equipment and storage medium |
CN110866655A (en) * | 2019-11-25 | 2020-03-06 | 武汉地铁运营有限公司 | Intelligent turnout jamming fault early warning method based on power numerical analysis |
WO2021109578A1 (en) * | 2019-12-02 | 2021-06-10 | 北京天元创新科技有限公司 | Method and apparatus for alarm prediction during service operation and maintenance, and electronic device |
WO2021121106A1 (en) * | 2019-12-20 | 2021-06-24 | 深圳前海微众银行股份有限公司 | Federated learning-based personalized recommendation method, apparatus and device, and medium |
CN111275503A (en) * | 2020-03-20 | 2020-06-12 | 京东数字科技控股有限公司 | Data processing method and device for acquiring lost user recall success rate |
CN111444428A (en) * | 2020-03-27 | 2020-07-24 | 腾讯科技(深圳)有限公司 | Information recommendation method and device based on artificial intelligence, electronic equipment and storage medium |
CN111432206A (en) * | 2020-04-24 | 2020-07-17 | 腾讯科技(北京)有限公司 | Video definition processing method and device based on artificial intelligence and electronic equipment |
CN111538908A (en) * | 2020-06-22 | 2020-08-14 | 腾讯科技(深圳)有限公司 | Search ranking method and device, computer equipment and storage medium |
CN111860971A (en) * | 2020-06-29 | 2020-10-30 | 交控科技股份有限公司 | Method and device for predicting residual life of turnout switch machine |
CN112036505A (en) * | 2020-09-17 | 2020-12-04 | 广西交控智维科技发展有限公司 | Method and device for determining equipment state of turnout switch machine and electronic equipment |
CN112150206A (en) * | 2020-09-28 | 2020-12-29 | 京东数字科技控股股份有限公司 | Method and equipment for predicting user interested article |
CN112477917A (en) * | 2020-11-10 | 2021-03-12 | 交控科技股份有限公司 | Turnout fault detection method and device, electronic equipment and storage medium |
CN112348170A (en) * | 2020-11-10 | 2021-02-09 | 交控科技股份有限公司 | Fault diagnosis method and system for turnout switch machine |
Non-Patent Citations (5)
Title |
---|
LEI XUE ET AL.: "Unsupervised anomaly detection system for railway turnout based on GAN", JOURNAL OF PHYSICS: CONFERENCE SERIES, 31 December 2019 (2019-12-31) * |
SHIZE HUANG ET AL.: "Turnout Fault Diagnosis through Dynamic Time Warping and Signal Normalization", JOURNAL OF ADVANCED TRANSPORTATION, 23 October 2017 (2017-10-23) * |
姬文江 等: "基于FastDTW的道岔故障智能诊断方法", 模式识别与人工智能, vol. 33, no. 11, 30 November 2020 (2020-11-30) * |
郝丽;李晓霞;: "缺陷汽车产品召回管理预警体系研究", 物流科技, no. 04, 10 April 2017 (2017-04-10) * |
黄蕾;: "基于道岔转辙机动作功率曲线关联分析道岔故障", 兵工自动化, no. 10, 15 October 2017 (2017-10-15) * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114692969A (en) * | 2022-03-29 | 2022-07-01 | 西门子交通技术(北京)有限公司 | Point switch failure prediction method, device, electronic equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108604360B (en) | Facility abnormity monitoring method and system | |
CN111178553A (en) | Industrial equipment health trend analysis method and system based on ARIMA and LSTM algorithms | |
Dong et al. | Equipment PHM using non-stationary segmental hidden semi-Markov model | |
US20060142990A1 (en) | Information processor, state judging unit and diagnostic unit, information processing method, state judging method and diagnosing method | |
CN104615122B (en) | A kind of industry control signal detection system and detection method | |
KR20210041724A (en) | Apparatus and method for predicting failure of electric car charger | |
KR101941854B1 (en) | System and method of estimating load with null data correction | |
CN110619587A (en) | Method and system for foundation pit monitoring intelligent early warning and data evidence storage | |
Wang et al. | Dynamic fault prognosis for multivariate degradation process | |
KR102373827B1 (en) | An Industrial Facility Operating Device Based on Standard Operation Level Evaluations and a Method for Operating It | |
CN113033772A (en) | Multi-equipment state monitoring method based on federal learning | |
CN113627496A (en) | Method, device, electronic equipment and readable storage medium for predicting fault of turnout switch machine | |
CN110058576A (en) | Equipment fault prognostics and health management method based on big data | |
CN117273440A (en) | Engineering construction Internet of things monitoring and managing system and method based on deep learning | |
CN107562041A (en) | Goat fault early warning method, device, equipment and computer-readable recording medium | |
CN114089206A (en) | Method, system, medium and device for predicting service life of battery redundancy module | |
Fernandez-Anakabe et al. | An Attribute Oriented Induction based Methodology for Data Driven Predictive Maintenance | |
Soualhi et al. | Dealing with prognostics uncertainties: Combination of direct and recursive remaining useful life estimations | |
Song et al. | Remaining Useful Life Prediction for Multi-state Stochastic Deterioration Assets Based on Phase-type Distributions | |
CN114186666B (en) | Generator coil temperature anomaly monitoring method based on self-standardized encoding and decoding | |
Shen et al. | Predictive maintenance on an elevator system using machine learning | |
Zhang et al. | Reliability evaluation of a disaster airflow emergency control system based on bayesian networks | |
US11639966B2 (en) | Enhanced electrical signature analysis for fault detection | |
CN117076260B (en) | Parameter and equipment abnormality detection method and device | |
RU2735296C1 (en) | Method for multi-parameter monitoring of complex electrotechnical objects state |
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 |