CN117408141A - Cable health state prediction and assessment method and device considering electric automobile access - Google Patents
Cable health state prediction and assessment method and device considering electric automobile access Download PDFInfo
- Publication number
- CN117408141A CN117408141A CN202311308022.XA CN202311308022A CN117408141A CN 117408141 A CN117408141 A CN 117408141A CN 202311308022 A CN202311308022 A CN 202311308022A CN 117408141 A CN117408141 A CN 117408141A
- Authority
- CN
- China
- Prior art keywords
- cable
- model
- cnn
- gpr
- bigru
- 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
- 230000036541 health Effects 0.000 title claims abstract description 72
- 238000000034 method Methods 0.000 title claims abstract description 52
- 238000001914 filtration Methods 0.000 claims abstract description 51
- 238000012544 monitoring process Methods 0.000 claims abstract description 32
- 238000012549 training Methods 0.000 claims abstract description 21
- 238000007781 pre-processing Methods 0.000 claims abstract description 8
- 238000005259 measurement Methods 0.000 claims description 11
- 230000008569 process Effects 0.000 claims description 9
- 238000013527 convolutional neural network Methods 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000005457 optimization Methods 0.000 claims description 7
- 238000011156 evaluation Methods 0.000 claims description 5
- 230000003862 health status Effects 0.000 claims description 5
- 230000002159 abnormal effect Effects 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims description 3
- 238000004140 cleaning Methods 0.000 claims description 2
- 238000010606 normalization Methods 0.000 claims description 2
- 238000012423 maintenance Methods 0.000 abstract description 6
- 241000272184 Falconiformes Species 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 230000000670 limiting effect Effects 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000000875 corresponding effect Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000009413 insulation Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000008439 repair process Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 241001465754 Metazoa Species 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000001010 compromised effect Effects 0.000 description 1
- 239000004020 conductor Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000036449 good health Effects 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000036961 partial effect Effects 0.000 description 1
- 230000005195 poor health Effects 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 238000013138 pruning Methods 0.000 description 1
- 230000002829 reductive effect Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
- 230000003313 weakening effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Abstract
The invention relates to a method and a device for predicting and evaluating the health state of a cable taking into account electric automobile access, comprising the following steps: acquiring monitoring data of cable equipment, and preprocessing the monitoring data; constructing a CNN-BiGRU model, optimizing the model by using an FHO algorithm, and training the optimized CNN-BiGRU model to obtain the FHO-CNN-BiGRU model; constructing a GPR filtering model, training the GPR filtering model, and obtaining a trained GPR filtering model; acquiring a cable equipment state prediction value through an FHO-CNN-BiGRU model and a trained GPR filtering model; the invention can help to predict potential faults by accurately evaluating the health state of the cable, and take corresponding maintenance measures, so that unexpected faults are avoided, the reliability and stability of a power supply system are improved, and the method is suitable for the requirement of large-scale access of electric automobiles.
Description
Technical Field
The invention belongs to the technical field of cable equipment monitoring, and particularly relates to a method and a device for predicting and evaluating the health state of cable equipment taking trolley bus access into consideration.
Background
With the popularization of electric vehicles and renewable energy sources, city planning construction and continuous development of communication technologies, power distribution equipment faces a plurality of challenges and changes. The cable is used as a key component in the power distribution equipment, is critical to safe and stable operation of the power distribution system, and compared with other equipment such as a circuit breaker, the cable has the characteristics of wide coverage range, large influence by humidity and the like, difficulty in comprehensive monitoring and the like, and the cable demand is increased year by year due to the rapid increase of load demands, so that the cable health state assessment and the risk and challenges of faults are continuously increased. Therefore, ensuring that the cable is maintained in a healthy state and early warning of faults is important to safe and stable operation of the power distribution system.
Charging large-scale electric vehicles brings higher load demands, and can lead to the cable bearing current exceeding rated load, thereby causing the problem of overload of the cable. In addition, the heat generated by the high load current also increases the temperature of the cable, potentially affecting its lifetime and presenting a potential safety risk. The factors make the health state evaluation and fault early warning of the cable under the condition that the electric automobile is connected into the power distribution network to be a problem to be solved urgently. Currently, health assessment and fault early warning studies on cables are mainly focused on the following aspects: firstly, working parameters of the cable, such as current, voltage, temperature and the like, are monitored in real time through a sensor and a monitoring system so as to acquire the operation state information of the cable. Secondly, the working state of the cable is diagnosed and evaluated by adopting the technologies of signal processing, data analysis, model establishment and the like so as to discover potential fault signs in advance. However, in the context of massive access of various novel elements to an actual power distribution system, existing researches and technologies have some shortcomings in the aspects of cable health state assessment and fault early warning, such as incomplete health state, to-be-improved early warning accuracy and the like.
Disclosure of Invention
The invention aims to provide a method and a device for predicting and evaluating the health state of a cable by considering the access of an electric automobile so as to improve the evaluation capability and the fault early warning level of the health state of the cable under the background of the large-scale access of the electric automobile to a power distribution network; by monitoring parameters such as working conditions, load states, temperature changes and the like of the cable in real time, the health state of the cable is evaluated, potential fault signs are found early and early warning is carried out, so that the service life of the cable is prolonged, a reliable and safe power supply system is finally provided for large-scale access of electric vehicles, and a new solution is provided for improving the reliability of a power distribution network and an electric vehicle charging station.
The invention solves the technical problems by adopting the following technical scheme:
the cable equipment health state prediction method considering trolley bus access comprises the following steps:
acquiring monitoring data of cable equipment, and preprocessing the monitoring data;
constructing a CNN-BiGRU model, optimizing the CNN-BiGRU model by using an FHO algorithm, and training the optimized CNN-BiGRU model to obtain the FHO-CNN-BiGRU model;
constructing a GPR filtering model, training the GPR filtering model, and obtaining a trained GPR filtering model;
and acquiring a cable equipment state predicted value through the CNN-BiGRU model and the trained GPR filtering model.
Further, the method for obtaining the cable equipment state predicted value through the CNN-BiGRU model and the trained GPR filtering model comprises the following steps:
and respectively inputting the preprocessed real-time monitoring data of the cable equipment into an FHO-CNN-BiGRU model and a trained GPR filtering model to obtain a cable equipment state predicted value and a GPR filtering result based on the FHO-CNN-BiGRU, inputting the obtained cable equipment state predicted value and the GPR filtering result into a GPR algorithm together to realize weighting, and finally obtaining the cable equipment state predicted value.
Further, the super parameters of the CNN-BiGRU model are optimized through an FHO algorithm, wherein the super parameters comprise the size of a convolution kernel, the number of hidden units of GRU and the learning rate.
Further, the method for optimizing the CNN-BiGRU model by using the FHO algorithm comprises the following steps:
step 1: defining a super parameter search space: determining the super parameters of the BIGRU model to be optimized, and defining a search range or a set of values for each super parameter;
step 2: initializing solution candidates: randomly generating a super-parameter combination of initial solution candidates according to the search space;
step 3: evaluating initial solution candidates;
step 4: setting a global optimal solution;
step 5: iterative optimization: the method comprises the steps that the fitness value of an updated super-parameter combination is evaluated, a global optimal solution is updated to be a current optimal solution, and an FHO algorithm gradually searches for a better super-parameter combination in an iterative optimization mode so as to improve the performance of a BiGRU model;
step 6: returning to the global best solution: and returning the global optimal solution to be used as the optimal super-parameter combination of the optimized BiGRU model after iteration is completed.
Further, in the CNN-biglu model, the convolutional neural network CNN performs feature extraction on the cable device monitoring data, captures key time sequence features, models the extracted time sequence features by the bi-directional gating circulation unit biglu to consider historical and future state information, uses the biglu hidden state of the last time step as a representation vector of the device health state, and inputs the representation vector to the full-connection layer for predicting the health state.
Further, the GPR filtering model firstly carries out offline training according to the nonlinear relation between the cable equipment history measurement data and the state data; and then obtaining a cable equipment state prediction result through the trained model according to the new measurement data.
The cable health state evaluation method considering the access of the electric automobile comprises the following steps:
according to the obtained cable equipment state predicted value, performing weight calculation on factors causing cable faults by using a analytic hierarchy process to obtain an ith fault factor weight w i ；
Obtaining real-time failure rate lambda of cable * Calculating the cable failure rate lambda (t) and the failure rate lambda after k overhauls k Correcting the fault rate by combining the service age of the cable to obtain the real-time fault rate of the cable:
λ * ＝w i ·λ k
wherein: k is the overhaul frequency of the cable equipment;
calculating real-time health index H of cable by using calculated real-time failure rate * ：
Wherein: c and K are curvature coefficient and proportionality coefficient respectively;
and (5) carrying out health state assessment on the cable according to the obtained real-time health index.
Further, according to the obtained real-time health index, the method for evaluating the health state of the cable comprises the following steps:
dividing the cable into different health states according to preset standards and thresholds, wherein the different health states comprise normal, early warning, abnormal and fault states;
and setting an alert threshold according to the real-time failure rate and the health index.
Consider cable health state prediction device that electric automobile inserted, include:
the data acquisition module is used for acquiring monitoring data of the cable equipment and preprocessing the monitoring data;
the FHO-CNN-BiGRU model acquisition module is used for constructing a CNN-BiGRU model, optimizing the model by using an FHO algorithm, and training the optimized CNN-BIGRU model to acquire the FHO-CNN-BiGRU model;
the GPR filtering model acquisition module is used for constructing a GPR filtering model, training the GPR filtering model and acquiring a trained GPR filtering model;
the cable equipment state prediction value acquisition module is used for respectively inputting the preprocessed real-time monitoring data of the cable equipment into the FHO-CNN-BiGRU model and the trained GPR filtering model to obtain a cable equipment state prediction value and a GPR filtering result based on the FHO-CNN-BiGRU, inputting the obtained cable equipment state prediction value and the obtained GPR filtering result into a GPR algorithm together to realize weighting, and finally obtaining the cable equipment state prediction value.
The invention has the advantages and positive effects that:
the cable health state prediction and assessment method suitable for large-scale electric automobile access has important significance for the research of cable health state assessment and fault early warning; firstly, the accurate assessment of the health state of the cable can help to predict potential faults and take corresponding maintenance measures to avoid occurrence of unexpected faults; secondly, the cable fault signs are found early and early warning is carried out, so that the maintenance cost and the maintenance time can be reduced, and the reliability and the stability of the power supply system are improved; in addition, the cable health state assessment and fault early warning are also beneficial to optimizing the operation of the power distribution network, improving the sustainability and manageability of the power distribution network, and adapting to the requirement of large-scale access of electric vehicles.
Drawings
The technical solution of the present invention will be described in further detail below with reference to the accompanying drawings and examples, but it should be understood that these drawings are designed for the purpose of illustration only and thus are not limiting the scope of the present invention. Moreover, unless specifically indicated otherwise, the drawings are intended to conceptually illustrate the structural configurations described herein and are not necessarily drawn to scale.
Fig. 1 is a flowchart of a method for predicting the health status of a cable device considering trolley access according to embodiment 1 of the present invention;
FIG. 2 is a structure of a Gaussian regression model according to embodiment 1 of the present invention;
fig. 3 is a flowchart of a method for evaluating the health status of cable equipment considering trolley access according to embodiment 1 of the present invention;
fig. 4 is a comparison chart of cable temperature changes before and after the electric automobile is connected in accordance with embodiment 2 of the present invention;
fig. 5 is a cable device information visualization interface based on the method for predicting the health status of cable devices according to embodiment 2 of the present invention;
fig. 6 is a cable monitoring real-time data display based on the method for predicting the health status of cable equipment according to embodiment 2 of the present invention.
Detailed Description
First, it should be noted that the following detailed description of the specific structure, characteristics, advantages, and the like of the present invention will be given by way of example, however, all descriptions are merely illustrative, and should not be construed as limiting the present invention in any way. Furthermore, any single feature described or implied in the embodiments mentioned herein, or any single feature shown or implied in the figures, may nevertheless be continued in any combination or pruning between these features (or equivalents thereof) to obtain still further embodiments of the invention that may not be directly mentioned herein.
It should be noted that, in the case of no conflict, the embodiments and features in the embodiments may be combined with each other.
Example 1
As shown in fig. 1, the method for predicting the health state of cable equipment considering trolley access provided in this embodiment includes the following steps:
acquiring historical monitoring data of cable equipment, and preprocessing the monitoring data;
constructing a CNN-BiGRU model, optimizing super parameters of the BiGRU model by using an FHO algorithm, and training the optimized CNN-BiGRU model according to historical data of cable equipment to obtain a FHO-CNN-BiGRU cable equipment state prediction model;
constructing a GPR filtering model, training the GPR filtering model according to historical data of the cable equipment, and obtaining a trained GPR filtering model;
and reading real-time monitoring data of the cable equipment, respectively inputting the FHO-CNN-BiGRU model and the trained GPR filtering model to obtain a cable equipment state predicted value and a GPR filtering result based on the FHO-CNN-BiGRU, inputting the obtained cable equipment state predicted value and the obtained GPR filtering result into the GPR weighting model together, and finally obtaining the cable equipment state predicted value.
After the monitoring data of the cable equipment is obtained, operations such as data cleaning and normalization are required, and data such as the temperature, the current-carrying capacity and the service life of the cable equipment are calculated according to the processed data so as to prepare input data.
Specific:
1) Cable temperature calculation model
The charging of the electric automobile can introduce nonlinear load, so that harmonic waves are generated; these harmonic signals have an effect on voltage and current, including harmonic voltage sigma u And harmonic current sigma i According to the IEC 61000 definition, respectively expressed as
Wherein: i 1 And I n The effective values of fundamental wave and n-order harmonic current are respectively; u (U) 1 And U n The effective values of fundamental wave and n-order harmonic voltage are respectively; h is the highest harmonic order.
According to IEC 60287 standard, 10kV three-core cable core temperature θ may be expressed as θ=θ c +θ 0 Wherein θ is c 、θ 0 The temperature rise and the environmental temperature of the cable are respectively;
wherein: r is R AC Is an alternating current resistor; t (T) 1 ～T 4 Respectively representing insulation layer thermal resistance, shielding layer to armor layer thermal resistance, armor layer and outer sheath thermal resistance and external thermal resistance; c i Is the core number of the cable lambda 2 The loss proportion of the armor layer is calculated;
2) Cable current-carrying capacity calculation model
Harmonic current in the electric automobile charging process can enable the total current carrying capacity of the cable to exceed a design value, overload and safety problems can be caused, and the current carrying capacity of the 10kV three-core cable can be calculated by the following formula:
wherein: lambda (lambda) 1 The loss of the shielding layer accounts for the conductor loss ratio;
3) Cable life calculation model
Harmonic current that produces in the electric automobile charging process can cause the extra heating of cable, produces the influence to the hot life-span of cable, and the cable life-span of per hour is:
wherein: l (L) h (T), wherein T (T) is the service life and the temperature of the cable in T hours; k (k) B And Δw are the Boltzmann constant and activation energy, respectively; l (L) 0 The service life of the cable is designed;
daily life loss L of cable d Represented as
In order to avoid the overload problem and the temperature rise phenomenon of the cable current-carrying capacity caused by the centralized charging of the electric automobile, the invention provides a cable equipment health state prediction method based on data driving and weighted Gaussian regression, which utilizes a convolutional neural network and a long-short-term memory network to predict parameters of the cable equipment to enhance the regression performance of the Gaussian process and is used for improving the accuracy of predicting the parameters such as the temperature, the current-carrying capacity and the like of the cable equipment.
According to the FHO-CNN-BiGRU-based cable equipment health state prediction method, a CNN algorithm is used for mining data relation and weakening noise interference, and then the data relation is imported into an FHO-based optimized BiGRU model for time sequence prediction, so that the purpose of improving prediction stability is achieved; and a Gaussian process regression method is adopted to carry out data smoothing by combining with a GPR filtering model, so that the accuracy and reliability of the data are improved.
In the invention, the super parameters of the CNN-BiGRU model are optimized through the FHO algorithm, wherein the super parameters comprise the convolution kernel size, the hidden unit number of GRU, the learning rate and other factors related to the model performance so as to improve the accuracy of the health state prediction.
The method for optimizing the CNN-BiGRU model by using the FHO algorithm comprises the following steps:
step 1, defining a super parameter search space: determining super parameters of a BiGRU model to be optimized, including learning rate, hidden layer unit number and the like, and determining the searching range of each super parameter;
step 2, generating and evaluating an initial solution: randomly initializing a group of initial solutions, namely initial super-parameters, training on a training data set according to the BiGRU model and the initial super-parameters, evaluating the performance of the model on a verification data set, and calculating the fitness value of each initial solution candidate through indexes such as accuracy, loss and the like;
step 3, setting the maximum iteration times and starting iterative optimization;
step 4, generating fire hawks and prey: in each iteration, an integer n is first randomly generated to determine the number of fire hawks, and then "fire hawks" and "prey" are determined in the search space;
step 5, calculating the distance between the fire hawk and the prey: calculating the total distance between the fire hawk and the prey in the search space, determining the hunting range of the fire hawk, and determining the territory of the fire hawk by dispersing the prey;
and 6, updating positions of fire hawks and prey: for each fire hawk, calculating the safe positions inside and outside the fire hawk territory, and then determining the new position of each hunting object;
step 7, evaluating the newly generated solution: training a BiGRU model by using the new super parameters, and calculating the fitness value of the newly generated solution;
and 8, updating a global optimal solution: in each iteration, comparing the advantages and disadvantages between the newly generated solution and the globally optimal solution, and if a better solution is found, updating the globally optimal solution;
step 9, checking termination conditions: judging whether the maximum iteration times are reached or whether the optimal solution is kept unchanged, and if so, ending the optimization process; otherwise, returning to the step 3 to continue optimization iteration;
step 10: returning to the global optimal solution: and returning the global optimal solution to serve as the optimal super-parameter combination of the optimized BiGRU model after iteration is completed.
In the CNN-BiGRU model, a convolutional neural network CNN performs feature extraction on cable equipment monitoring data, captures key time sequence features, models the extracted time sequence features by a bi-directional gating circulating unit BiGRU to consider historical and future state information, uses a BiGRU hidden state of the last time step as a representation vector of the equipment health state, and inputs the representation vector into a full-connection layer for predicting the health state.
Firstly, performing offline training on a GPR filtering model according to a nonlinear relation between cable equipment historical measurement data and state data; and then obtaining a cable equipment state prediction result through the trained model according to the new measurement data.
Combining with an FHO-CNN-BiGRU prediction model, the invention utilizes the superiority of a GPR model in processing nonlinear problems to obtain nonlinear mapping relation between quantity measurement and state quantity, then inputs the newly obtained quantity measurement into the model, and obtains an expected state estimation result through calculation;
the Gaussian regression model maps the input features to a high-dimensional space through a set of basis functions psi (, so that a linear relation between data is found in the high-dimensional space, and the quantity measurement Z is represented by the basis functions psi (x), so that the probability of new data can be obtained.
The confidence coefficient is adopted for regression in the Gaussian process to judge the estimation result, and the estimation method can weaken noise interference, so that the accuracy of the estimation result is improved; the Gaussian regression model structure is shown in FIG. 2, and offline training is performed according to the nonlinear relation between the historical measurement data and the state data of the cable equipment; and then obtaining a cable equipment state prediction result through the trained model according to the new measurement data.
The cable health state evaluation method considering the access of the electric automobile comprises the following steps:
according to the obtained cable equipment state predicted value, performing weight calculation on factors causing cable faults by using a analytic hierarchy process to obtain an ith fault factor weight w i The method comprises the steps of carrying out a first treatment on the surface of the Among these, influencing factors include: heavy overload, lightning strike, high temperature, moisture, chemical corrosion, human damage, animal damage, and insulation aging;
obtaining real-time failure rate lambda of cable * Calculating the cable failure rate lambda (t) and the failure rate lambda after k times of overhauling through a formula (11) k And correcting the fault rate by adopting a formula (12) in combination with the service age of the cable to obtain the real-time fault rate of the cable.
λ * ＝w i ·λ k (12)
Wherein: k is the overhaul frequency of the cable equipment;
calculating real-time health index H of cable by using calculated real-time failure rate * ：
Wherein: c and K are curvature coefficient and proportionality coefficient respectively;
and (5) carrying out health state assessment on the cable according to the obtained real-time health index. Specific: the cable can be divided into different health states according to preset standards and thresholds, including normal, early warning, abnormal and fault states.
Normal (80-100): the cable is in good condition with no obvious signs of failure. The range indicates that the cable is in good health state and runs stably.
Early warning (60-80): cables show some potential signs of problems or failure, but have not yet reached severity; this range indicates that there is a certain risk of the health of the cable, which may need to be noted and monitored.
Abnormality (30 to 60): obvious faults or problems exist in the cable, and the cable can cause performance degradation or partial functional failure; this range indicates that the cable is in poor health and requires repair or maintenance.
Fault (0-30): the cable has failed and is not functioning properly or providing the desired function. This range indicates that the health of the cable is severely compromised and needs urgent repair or replacement.
According to the real-time failure rate and the health index, an alert threshold can be set; when the failure rate exceeds a threshold or the health index falls, the system may issue an alarm, alerting maintenance personnel to take corresponding action, such as repairing, replacing or upgrading the cable.
Example 2
In the embodiment, a test is performed by taking an actual cable network in a male security area as an example to verify the monitoring effect of the proposed cable equipment health state prediction method on cable faults. Fig. 4 compares the temperature change diagrams of the cable after the electric automobile is connected with the cable before the electric automobile is connected, meanwhile, fig. 5 gives detailed information such as the health state of the cable and an alarm, the temperature of the cable is obviously increased in the charging load peak period, the influence on the cable equipment after the electric automobile is connected with the load of a large scale of electric automobile can be accurately monitored by the cable health state monitoring method provided by the invention, and an early warning can be given when the temperature exceeds a threshold value. Fig. 6 shows the results of temperature, discharge capacity and the like after monitoring by adopting the method, and the display results show that the method for predicting the health state of the cable taking into consideration the electric automobile access can monitor the information of the highest temperature, the lowest temperature, the discharge capacity and the like of the cable in real time and visualize the information, is beneficial to early finding out the fault signs of the cable and early warning, and improves the reliability and the stability of a power supply system.
Example 3
Based on the same inventive concept, the embodiment of the application also provides a cable health state prediction device considering electric automobile access, comprising:
the data acquisition module is used for acquiring monitoring data of the cable equipment and preprocessing the monitoring data;
the FHO-CNN-BiGRU model acquisition module is used for constructing a CNN-BiGRU model, optimizing the model by using an FHO algorithm, and training the optimized CNN-BiGRU model to acquire the FHO-CNN-BiGRU model;
the GPR filtering model acquisition module is used for constructing a GPR filtering model, training the GPR filtering model and acquiring a trained GPR filtering model;
the cable equipment state prediction value acquisition module is used for respectively inputting the preprocessed real-time monitoring data of the cable equipment into the FHO-CNN-BiGRU model and the trained GPR filtering model to obtain a cable equipment state prediction value and a GPR filtering result based on the FHO-CNN-BiGRU, inputting the obtained cable equipment state prediction value and the obtained GPR filtering result into a GPR algorithm together to realize weighting, and finally obtaining the cable equipment state prediction value.
The foregoing examples illustrate the invention in detail, but are merely preferred embodiments of the invention and are not to be construed as limiting the scope of the invention. All equivalent changes and modifications within the scope of the present invention are intended to be covered by the present invention.
Claims (11)
1. The cable health state prediction method considering electric automobile access is characterized by comprising the following steps:
acquiring monitoring data of cable equipment, and preprocessing the monitoring data;
constructing a CNN-BiGRU model, optimizing the CNN-BiGRU model by using an FHO algorithm, and training the optimized CNN-BiGRU model to obtain the FHO-CNN-BiGRU model;
constructing a GPR filtering model, training the GPR filtering model, and obtaining a trained GPR filtering model;
and acquiring a cable equipment state prediction value through the FHO-CNN-BiGRU model and the trained GPR filtering model.
2. The method for predicting the state of health of a cable by considering electric vehicle access according to claim 1, wherein the method for obtaining the predicted value of the state of the cable equipment by using the FHO-CNN-BiGRU model and the trained GPR filtering model is as follows:
and respectively inputting the preprocessed real-time monitoring data of the cable equipment into an FHO-CNN-BiGRU model and a trained GPR filtering model to obtain a cable equipment state predicted value and a GPR filtering result based on the FHO-CNN-BiGRU, inputting the obtained cable equipment state predicted value and the GPR filtering result into a GPR algorithm together to realize weighting, and finally obtaining the cable equipment state predicted value.
3. The method for predicting the health of the cable taking into account electric automobile access according to claim 1, wherein the super parameters of the CNN-BiGRU model are optimized through FHO algorithm, and the super parameters comprise convolution kernel size, hidden unit number of GRU and learning rate.
4. The method for predicting the health state of a cable taking into account electric vehicle access according to claim 1, wherein the method for optimizing the CNN-biglu model by using the FHO algorithm is as follows:
step 1: defining a super parameter search space: determining the super parameters of the BiGRU model to be optimized, and defining a search range or a set of values for each super parameter;
step 2: initializing solution candidates: randomly generating a super-parameter combination of initial solution candidates according to the search space;
step 3: evaluating initial solution candidates;
step 4: setting a global optimal solution;
step 5: iterative optimization: the method comprises the steps that the fitness value of an updated super-parameter combination is evaluated, a global optimal solution is updated to be a current optimal solution, and an FHO algorithm gradually searches for a better super-parameter combination in an iterative optimization mode so as to improve the performance of a BiGRU model;
step 6: returning to the global best solution: and returning the global optimal solution to be used as the optimal super-parameter combination of the optimized BiGRU model after iteration is completed.
5. The method for predicting the health of a cable taking into account electric vehicle access according to claim 1, wherein: in the CNN-BiGRU model, a convolutional neural network CNN performs feature extraction on cable equipment monitoring data, captures key time sequence features, models the extracted time sequence features by a bi-directional gating circulating unit BiGRU to consider historical and future state information, uses a BiGRU hidden state of the last time step as a representation vector of the equipment health state, and inputs the representation vector into a full-connection layer for predicting the health state.
6. The method for predicting the health of a cable taking into account electric vehicle access according to claim 1, wherein: firstly, performing offline training on a GPR filtering model according to a nonlinear relation between cable equipment historical measurement data and state data; and then obtaining a cable equipment state prediction result through the trained model according to the new measurement data.
7. A cable health state evaluation method based on the cable health state prediction method considering electric vehicle access according to any one of claims 1 to 6, characterized by comprising the steps of:
according to the obtained cable equipment state predicted value, performing weight calculation on factors causing cable faults by using a analytic hierarchy process to obtain an ith fault factor weight w i ；
Obtaining real-time failure rate lambda of cable * ；
Calculating real-time health index H of cable by using calculated real-time failure rate of cable * ；
And (5) carrying out health state assessment on the cable according to the obtained real-time health index.
8. The method for evaluating the health state of the cable taking into account electric vehicle access according to claim 7, wherein the method for evaluating the health state of the cable according to the obtained real-time health index is as follows:
dividing the cable into different health states according to preset standards and thresholds, wherein the different health states comprise normal, early warning, abnormal and fault states;
and setting an alert threshold according to the real-time failure rate and the health index of the cable.
9. The method for evaluating the health status of a cable taking into account electric vehicle access as recited in claim 7, wherein the real-time failure rate λ of the cable * The acquisition method of (1) comprises the following steps:
calculating the cable failure rate lambda (t) and the failure rate lambda after k times of overhauling k Correcting the fault rate by combining the service age of the cable to obtain the real-time fault rate of the cable:
λ * ＝w i ·λ k (12)
wherein: k is the overhaul frequency of the cable equipment;
calculating real-time health index H of cable by using calculated real-time failure rate * The method of (1) is as follows:
wherein: c and K are the curvature coefficient and the proportionality coefficient, respectively.
10. The method for evaluating the health state of a cable taking into account electric vehicle access according to claim 7, wherein after the monitored data of the cable equipment are acquired, data cleaning and normalization preprocessing are required, and cable temperature, cable current-carrying capacity and cable service life are calculated according to the preprocessed data.
11. Consider cable health state prediction device that electric automobile inserted, its characterized in that includes:
the data acquisition module is used for acquiring monitoring data of the cable equipment and preprocessing the monitoring data;
the FHO-CNN-BiGRU model acquisition module is used for constructing a CNN-BiGRU model, optimizing the model by using an FHO algorithm, and training the optimized CNN-BiGRU model to acquire the FHO-CNN-BiGRU model;
the GPR filtering model acquisition module is used for constructing a GPR filtering model, training the GPR filtering model and acquiring a trained GPR filtering model;
the cable equipment state prediction value acquisition module is used for respectively inputting the preprocessed real-time monitoring data of the cable equipment into the FHO-CNN-BiGRU model and the trained GPR filtering model to obtain a cable equipment state prediction value and a GPR filtering result based on the FHO-CNN-BiGRU, inputting the obtained cable equipment state prediction value and the obtained GPR filtering result into a GPR algorithm together to realize weighting, and finally obtaining the cable equipment state prediction value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311308022.XA CN117408141A (en) | 2023-10-10 | 2023-10-10 | Cable health state prediction and assessment method and device considering electric automobile access |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311308022.XA CN117408141A (en) | 2023-10-10 | 2023-10-10 | Cable health state prediction and assessment method and device considering electric automobile access |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117408141A true CN117408141A (en) | 2024-01-16 |
Family
ID=89491804
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311308022.XA Pending CN117408141A (en) | 2023-10-10 | 2023-10-10 | Cable health state prediction and assessment method and device considering electric automobile access |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117408141A (en) |
-
2023
- 2023-10-10 CN CN202311308022.XA patent/CN117408141A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105866638B (en) | A kind of city net cable connector state of insulation on-line monitoring prior-warning device and method | |
| CN111965476B (en) | Low-voltage diagnosis method based on graph convolution neural network | |
| Li et al. | Fault diagnosis for lithium-ion batteries in electric vehicles based on signal decomposition and two-dimensional feature clustering | |
| CN108445410A (en) | A kind of method and device of monitoring accumulator group operating status | |
| CN109583520B (en) | State evaluation method of cloud model and genetic algorithm optimization support vector machine | |
| CN116614177B (en) | Optical fiber state multidimensional parameter monitoring system | |
| CN115979349A (en) | Power station space environment monitoring method and system | |
| CN113987294A (en) | CVT (continuously variable transmission) online fault diagnosis method based on genetic optimization GRU (generalized regression Unit) neural network | |
| CN111856209A (en) | Power transmission line fault classification method and device | |
| CN107292512B (en) | Power equipment space-time multi-dimensional safety assessment method based on symbolic dynamics and hidden Markov model | |
| CN115271238A (en) | Method and device for predicting service life of transformer | |
| CN114997566A (en) | Power grid blocking risk assessment method and system considering node connectivity loss | |
| Mharakurwa | In-service power transformer life time prospects: review and prospects | |
| CN117408141A (en) | Cable health state prediction and assessment method and device considering electric automobile access | |
| CN109038550B (en) | Power system self-repeatability index calculation method based on voltage static stability | |
| CN116070140A (en) | Power distribution substation safe operation state monitoring system and method | |
| CN113933718B (en) | Retired battery capacity sorting method, retired battery capacity sorting device, retired battery capacity sorting equipment and storage medium | |
| CN111060755A (en) | Electromagnetic interference diagnosis method and device | |
| CN115859594A (en) | Health evaluation algorithm of power transformation equipment based on hierarchical analysis | |
| CN115017828A (en) | Power cable fault identification method and system based on bidirectional long-short-time memory network | |
| CN114879070A (en) | Battery state evaluation method and related equipment | |
| Negri et al. | Artificial Intelligence and Smart Grids: the Cable Joint Test Case | |
| CN113640596A (en) | Converter transformer abnormity detection method and device, computer equipment and storage medium | |
| CN111337764A (en) | Charging pile fault diagnosis system and method and storage medium | |
| CN112782603A (en) | Lithium ion battery cycle life distribution fitting method based on interval truncation data |
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 |