CN111546854B - On-road identification and diagnosis method for intelligent train air conditioning unit - Google Patents
On-road identification and diagnosis method for intelligent train air conditioning unit Download PDFInfo
- Publication number
- CN111546854B CN111546854B CN202010558477.7A CN202010558477A CN111546854B CN 111546854 B CN111546854 B CN 111546854B CN 202010558477 A CN202010558477 A CN 202010558477A CN 111546854 B CN111546854 B CN 111546854B
- Authority
- CN
- China
- Prior art keywords
- parameters
- normal
- conditioning unit
- air conditioning
- vector
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60H—ARRANGEMENTS OF HEATING, COOLING, VENTILATING OR OTHER AIR-TREATING DEVICES SPECIALLY ADAPTED FOR PASSENGER OR GOODS SPACES OF VEHICLES
- B60H1/00—Heating, cooling or ventilating [HVAC] devices
- B60H1/00507—Details, e.g. mounting arrangements, desaeration devices
- B60H1/00585—Means for monitoring, testing or servicing the air-conditioning
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B61—RAILWAYS
- B61D—BODY DETAILS OR KINDS OF RAILWAY VEHICLES
- B61D27/00—Heating, cooling, ventilating, or air-conditioning
- B61D27/0018—Air-conditioning means, i.e. combining at least two of the following ways of treating or supplying air, namely heating, cooling or ventilating
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Abstract
The invention discloses an on-the-way identification and diagnosis method for an intelligent train air conditioning unit, wherein the method comprises the following steps: acquiring internal operation parameters of the train air conditioning unit respectively operating under various preset environmental parameters and according to normal working conditions, and constructing a normal parameter database; acquiring internal operation parameters of the train air conditioning unit respectively under various preset environmental parameters and different fault working conditions, and constructing an abnormal parameter database; finding normal parameters matched with each group of abnormal parameters and normal parameters in a normal parameter database, and calculating corresponding residual errors to serve as training samples; training to obtain a fault prediction model according to all training samples and corresponding working condition types; and obtaining a real-time residual error time sequence by the same method, inputting the real-time residual error time sequence to the fault prediction model to obtain a plurality of working condition type prediction results, and comprehensively evaluating to obtain a final real-time working condition type. The invention can improve the failure prediction precision.
Description
Technical Field
The invention relates to the field of train air conditioning unit fault identification, in particular to an on-the-way identification and diagnosis method for an intelligent train air conditioning unit.
Background
With the rapid development of passenger trains in modern society, the application of train air conditioning units on passenger trains is becoming more extensive. The current passenger train has better sealing performance, which directly causes poor air circulation in the carriage; in spring peak hours, the density of people in the train is higher, the requirement on ventilation in the carriage is higher, and therefore the train air conditioner is required to work for a long time without faults. Therefore, the method for accurately diagnosing the faults of the train air conditioning unit has very important significance.
The air-conditioning refrigeration system comprises main elements such as an evaporator, a condenser, a compressor and the like, and all the elements cooperate with each other to jointly complete the refrigeration function of the air conditioner. The basic operation principle of the refrigeration cycle is shown in fig. 1. The types of air conditioning faults can be classified according to the types of air conditioning unit components as follows: refrigerant leakage, evaporating dish blockage, condenser scaling, non-condensable gas and compressor shutdown.
At present, most of train air conditioning unit fault diagnosis methods are to directly establish a real-time mapping model of physical characteristics and fault types to predict train air conditioning unit faults by measuring the physical characteristics of all components of the air conditioning unit, namely, fault detection is carried out on the train air conditioning unit at a certain moment through the physical characteristics at the moment, but as sensors possibly have sudden jumping and the number of the sensors is large, the false detection rate of the diagnosis method is high.
In addition, the brands and mechanical structures of different train air conditioning units are different, so that the absolute value of the physical characteristic is not reasonable as a fault diagnosis standard, and the fault prediction accuracy is reduced.
Disclosure of Invention
The invention provides an on-the-way identification and diagnosis method for an intelligent train air conditioning unit, which can solve the problem of high false detection rate caused by sudden jumping of a sensor.
In order to achieve the technical purpose, the invention adopts the following technical scheme:
an on-the-way identification and diagnosis method of an intelligent train air conditioning unit based on physical characteristic residual errors comprises the following steps:
step 2, obtaining internal operation parameters of the train air conditioning unit when the train air conditioning unit operates under various preset environmental parameters and different fault working conditions; taking internal operation parameters of each different environment parameter and different fault working conditions in operation together with corresponding environment parameters as a group of abnormal parameters; the same environmental parameters and the same fault working conditions correspondingly acquire a plurality of groups of abnormal parameters, and all the abnormal parameters form an abnormal parameter database;
step 3, traversing each group of abnormal parameters and normal parameters, finding normal parameters matched with the environmental parameters in a normal parameter database according to the principle of minimum similarity, and performing difference calculation on internal operation parameters in the normal parameters matched with the corresponding environmental parameters of each group of abnormal parameters and normal parameters respectively to obtain corresponding residual errors;
residual errors corresponding to each group of abnormal parameters and normal parameters are used as a training sample;
step 4, training an extreme learning machine model by taking all training samples as input and taking corresponding working condition types as output to obtain a fault prediction model of the train air conditioning unit;
step 5, acquiring an environmental parameter time sequence of the train air conditioning unit within a preset duration and a corresponding internal operation parameter time sequence in real time at a preset frequency, and acquiring a corresponding residual error time sequence according to the step 3; and then inputting the residual error time sequence into a fault prediction model of the train air conditioning unit, and taking one working condition type with the largest proportion in all output working condition type prediction results, namely the real-time working condition type of the train air conditioning unit.
In a more preferred technical solution, the finding of the normal parameter matching the environmental parameter in the normal parameter database according to the principle of minimum similarity specifically includes:
for each group of abnormal parameters to be matched: calculating Euclidean distances between each group of normal parameters in the normal parameter database and the current abnormal parameters to be matched, and taking a group of normal parameters with the minimum Euclidean distance as a matching case of the current abnormal parameters to be matched, namely a group of normal parameters matched with environmental parameters in the current abnormal parameters to be matched;
for each set of normal parameters to be matched: and calculating Euclidean distances between other groups of normal parameters except the normal parameters and the current normal parameters to be matched in the normal parameter database, and taking a group of normal parameters with the minimum Euclidean distance as a matching case of the current normal parameters to be matched, namely a group of normal parameters matched with the environmental parameters in the current normal parameters to be matched.
In a more preferred embodiment, the types of the operating conditions for training the extreme learning machine model include: normal operating mode, refrigerant leak, evaporating dish jam, condenser scale deposit, have non-condensable gas and compressor to shut down, other operating mode types except normal operating mode all belong to the fault condition.
In a more preferred technical solution, the preset frequency in step 5 is to obtain 20 groups of data per minute, and the preset time is 2.5 minutes.
In a more preferred embodiment, the environmental parameters include: setting temperature of the air conditioning unit, inlet air dry bulb temperature of the air conditioning unit, inlet air relative humidity of the air conditioning unit and outside air dry bulb temperature; the internal operating parameters include: the air suction temperature of the compressor, the air exhaust temperature of the compressor, the air suction pressure of the compressor, the exhaust pressure of the compressor and the refrigerating capacity of the air conditioner of the train.
In a more preferred technical solution, the euclidean distance calculating method includes:
wherein the content of the first and second substances,
tisetting temperature, air conditioning unit inlet air dry bulb temperature, air conditioning unit inlet air relative humidity and outside air dry bulb temperature for the ith group of normal parameters in the normal parameter database respectively; KT (karat)set、KTin、KWinAnd T is respectively set temperature of the air conditioning unit, inlet air dry bulb temperature of the air conditioning unit, inlet air relative humidity of the air conditioning unit and outside air dry bulb temperature in the abnormal parameters to be matched, or is respectively set temperature of the air conditioning unit, inlet air dry bulb temperature of the air conditioning unit, inlet air relative humidity of the air conditioning unit and outside air dry bulb temperature in the normal parameters to be matched.
In a more preferred technical scheme, the parameters of the extreme learning machine optimized by adopting a differential evolution algorithm comprise an input weight vector and a hidden layer neuron bias vector, and the optimization process comprises the following steps:
step 4.1, parameter setting and population initialization: setting maximum evolution algebra GmaxPopulation size NP, scaling factor F and crossover probability Pcr(ii) a Representing each individual in the population as a target vector r ═ W b consisting of the input weight vector and the hidden layer neuron bias vector]For the input weight vector of the parameter vector, each element is in the range [0, 1 ]]Performing random initialization on each element in hidden layer neuron bias vector of parameter vector in range of [ -5, 5]Carrying out random initialization; taking the variance between the predicted value and the true value of the extreme learning machine as a fitness function;
step 4.2, making the evolution algebra G equal to 0; the target vector r is initialized to p (0);
step 4.3, making the evolution algebra G equal to G + 1; p (G +1) ═ p (G);
step 4.4, randomly selecting 3 individuals r1, r2 and r3 in the population outside the target vector p (G);
step 4.5, carrying out differential variation operation to generate a variation vector;
r*(G)＝r1(G)+F*(r2(G)-r3(G))
wherein F is called a scaling factor and is a constant; r is*Is a variation vector;
step 4.6, performing cross operation on the target vector and the variation vector to generate a test vector;
specifically, each component of the test vector is subjected to a crossover operation:
wherein the content of the first and second substances,
is the G-th generation test vector component; pcrIs the cross probability, is a constant;
is the G generation variation vector component;
is a component of the G-th generation target vector.
4.7, calculating the fitness value of the test vector, and performing comparison selection operation;
respectively substituting the test vector and the target vector into an extreme learning machine to calculate corresponding fitness function values, and taking the fitness function value which is more optimal as a new target vector;
wherein, r (G) is the G-th generation target vector; and f is a fitness function.
Step 4.8, if G ═ GmaxAnd ending the optimization process and outputting the network weight W of the current target vector and the hidden layer neuron offset b, otherwise, turning to the step 4.3.
Advantageous effects
When the fault prediction of the train air conditioning unit is carried out, the normal parameters matched with the environmental parameters are selected, so that the difference between the internal operation parameters obtained in real time and the internal operation parameters in the normal parameters matched with the internal operation parameters is further carried out to obtain a residual error, namely a physical characteristic relative value comprising fault characteristics, and the fault prediction is carried out by utilizing the trained fault prediction model of the train air conditioning unit, so that the fault prediction error judgment of environmental factors on the fault prediction can be avoided, and the influence of the absolute value of the physical characteristic on the fault prediction precision of different types of air conditioning units is avoided;
in addition, multiple groups of environmental parameters and internal operation parameter data are obtained in real time to further obtain multiple fault prediction results, so that the working condition type of the air conditioning unit is comprehensively evaluated, the interference of sensor sudden jumping on fault prediction is reduced to a great extent, and the fault prediction precision is improved.
Drawings
FIG. 1 is a basic working schematic diagram of a refrigeration cycle of an air conditioning unit of a train;
FIG. 2 is a diagram illustrating a method according to an embodiment of the present invention for finding a matching case for a current abnormal parameter to be matched in a normal parameter database;
FIG. 3 is a block diagram of an extreme learning machine model;
fig. 4 is a schematic flow chart of the method according to the embodiment of the present invention.
Detailed Description
The following describes embodiments of the present invention in detail, which are developed based on the technical solutions of the present invention, and give detailed implementation manners and specific operation procedures to further explain the technical solutions of the present invention.
The embodiment provides an on-the-way identification and diagnosis method for an intelligent train air conditioning unit, which includes the steps of firstly preprocessing historical data to obtain a training sample, then training to obtain a train air conditioning unit fault prediction model, and further predicting faults of the train air conditioning unit, as shown in fig. 4, specifically including the following steps:
the environmental parameters include: train air conditioning unit settlement temperature, unit import air dry bulb temperature, outside air dry bulb temperature, unit import air relative humidity, measured inside operating parameter includes: the air suction temperature of the compressor, the air exhaust temperature of the compressor, the air suction pressure of the compressor, the exhaust pressure of the compressor and the refrigerating capacity of the air conditioner of the train. The above environmental parameters and the internal operation parameters are respectively measured by installing corresponding sensors inside and outside each part of the train air conditioning unit to obtain historical data, and at least one parameter value is different among the four environmental parameters, namely the set temperature of the train air conditioning unit, the dry bulb temperature of the inlet air of the unit, the dry bulb temperature of the outside air and the relative humidity of the inlet air of the unit, among the multiple preset environmental parameters.
Step 2, obtaining internal operation parameters of the train air conditioning unit when the train air conditioning unit operates under various preset environmental parameters and different fault working conditions; taking internal operation parameters of each different environment parameter and different fault working conditions in operation together with corresponding environment parameters as a group of abnormal parameters; the same environmental parameters and the same fault working conditions correspondingly acquire a plurality of groups of abnormal parameters, and all the abnormal parameters form an abnormal parameter database;
the method for acquiring the abnormal parameters in the step 2 is the same as the method for acquiring the normal parameters in the step 1, and the difference is only that the data in the step is acquired when the data is operated under various fault conditions.
The working condition types in the embodiment include the normal working condition in the step 1 and the fault working condition in the step 2, and the fault working condition is further divided into four working condition types of refrigerant leakage, evaporating dish blockage, condenser scaling, non-condensable gas and compressor shutdown.
Step 3, traversing each group of abnormal parameters and normal parameters, finding normal parameters matched with the environmental parameters in a normal parameter database according to the principle of minimum similarity, and performing difference calculation on internal operation parameters in the normal parameters matched with the corresponding environmental parameters of each group of abnormal parameters and normal parameters respectively to obtain corresponding residual errors; residual errors corresponding to each group of abnormal parameters and normal parameters are used as a training sample;
the method comprises the following steps of finding a normal parameter matched with an environmental parameter in a normal parameter database according to the principle of minimum similarity, specifically:
as shown in fig. 2, for each set of anomaly parameters to be matched: calculating Euclidean distances between each group of normal parameters in the normal parameter database and the current abnormal parameters to be matched, and taking a group of normal parameters with the minimum Euclidean distance as a matching case of the current abnormal parameters to be matched, namely a group of normal parameters matched with environmental parameters in the current abnormal parameters to be matched;
as shown in fig. 2, for each set of normal parameters to be matched: and calculating Euclidean distances between other groups of normal parameters except the normal parameters and the current normal parameters to be matched in the normal parameter database, and taking a group of normal parameters with the minimum Euclidean distance as a matching case of the current normal parameters to be matched, namely a group of normal parameters matched with the environmental parameters in the current normal parameters to be matched.
The Euclidean distance calculation formula is as follows:
wherein the content of the first and second substances,
tisetting temperature, air conditioning unit inlet air dry bulb temperature, air conditioning unit inlet air relative humidity and outside air dry bulb temperature for the ith group of normal parameters in the normal parameter database respectively; KT (karat)set、KTin、KWinAnd T is respectively set temperature of the air conditioning unit, inlet air dry bulb temperature of the air conditioning unit, inlet air relative humidity of the air conditioning unit and outside air dry bulb temperature in the abnormal parameters to be matched, or is respectively set temperature of the air conditioning unit, inlet air dry bulb temperature of the air conditioning unit, inlet air relative humidity of the air conditioning unit and outside air dry bulb temperature in the normal parameters to be matched.
After obtaining a matching case with the current abnormal parameters or normal parameters to be matched, performing difference calculation on internal operation parameters in each group of abnormal parameters (namely the current abnormal parameters to be matched) and normal parameters (namely the current normal parameters to be matched) respectively matched with the corresponding environmental parameters (namely the matching case), and obtaining corresponding residual errors as follows:
ΔP＝|P-Pi*|；
wherein the content of the first and second substances,
pi*respectively matching the compressor air suction temperature, the compressor air exhaust temperature, the compressor air suction pressure, the compressor exhaust pressure and the train air conditioner refrigerating capacity of the case; YTin、YTout、YPin、YPoutAnd P is the compressor air suction temperature, the compressor air exhaust temperature, the compressor air suction pressure, the compressor exhaust pressure and the train air conditioner refrigerating capacity of the current abnormal parameter or normal parameter to be matched respectively; Δ YTin、ΔYTout、ΔYPin、ΔYPoutAnd delta P is respectively a compressor air suction temperature residual error, a compressor air exhaust temperature residual error, a compressor air suction pressure residual error, a compressor exhaust pressure residual error and a train air conditioner refrigerating capacity residual error corresponding to the current abnormal parameter or normal parameter to be matched.
Step 4, training an extreme learning machine model by taking all training samples as input and taking corresponding working condition types as output to obtain a fault prediction model of the train air conditioning unit;
in training the extreme learning machine model, the labels for the operating condition types are set as: the normal working condition is 0, the refrigerant leakage is 1, the blockage of an evaporating dish is 2, the scaling of a condenser is 3, the non-condensable gas is 4, and the shutdown of a compressor is 5.
Because the extreme learning machine only needs to set the number of hidden layer nodes of the network, parameters of the extreme learning machine do not need to be adjusted in the training process, and a unique optimal solution is generated, the extreme learning machine is selected to train to obtain the train unit fault prediction model, and the method has the advantages of high learning speed and good generalization performance. The structure of the limit learning machine is shown in fig. 3, and the limit learning machine parameters are set as follows in the present embodiment: the number of input neurons of the extreme learning machine is 5, the number N of neurons of an implicit layer is determined according to experience, the number of output neurons is 1, and the number of iterations is 100; and the input weight W and the hidden layer neuron bias b are initially assigned randomly and then optimized by a differential evolution algorithm.
In this embodiment, for training of the extreme learning machine model, parameters of the extreme learning machine are specifically optimized by using a differential evolution algorithm, that is, a weight vector and a hidden layer neuron bias vector are input, and the specific process is as follows:
step 4.1, parameter setting and population initialization: setting maximum evolution algebra GmaxPopulation size NP, scaling factor F and crossover probability Pcr(ii) a Representing each individual in the population as a target vector r ═ W b consisting of the input weight vector and the hidden layer neuron bias vector]For the input weight vector of the parameter vector, each element is in the range [0, 1 ]]Performing random initialization on each element in hidden layer neuron bias vector of parameter vector in range of [ -5, 5]Carrying out random initialization; taking the variance between the predicted value and the true value of the extreme learning machine as a fitness function;
step 4.2, making the evolution algebra G equal to 0; the target vector r is initialized to p (0);
step 4.3, let evolution generation G ═ G +1, p (G +1) ═ p (G);
step 4.4, randomly selecting 3 individuals r1, r2 and r3 in the population outside the target vector p (G);
step 4.5, carrying out differential variation operation to generate a variation vector;
r*(G)＝r1(G)+F*(r2(G)-r3(G))
wherein F is called a scaling factor and is a constant; r is*Is a variation vector;
step 4.6, performing cross operation on the target vector and the variation vector to generate a test vector;
specifically, each component of the test vector is subjected to a crossover operation:
wherein the content of the first and second substances,
is the G-th generation test vector component; pcrIs the cross probability, is a constant;
is the G generation variation vector component;
is a component of the G-th generation target vector.
4.7, calculating the fitness value of the test vector, and performing comparison selection operation;
respectively substituting the test vector and the target vector into an extreme learning machine to calculate corresponding fitness function values, and taking the fitness function value which is more optimal as a new target vector;
wherein, r (G) is the G-th generation target vector; and f is a fitness function.
Step 4.8, if G ═ GmaxAnd ending the optimization process and outputting the network weight W of the current target vector and the hidden layer neuron offset b, otherwise, turning to the step 4.3.
Step 5, acquiring an environmental parameter time sequence of the train air conditioning unit within a preset duration and a corresponding internal operation parameter time sequence in real time at a preset frequency, and acquiring a corresponding residual error time sequence according to the step 3; and then inputting the residual error time sequence into a fault prediction model of the train air conditioning unit, and taking one working condition type with the largest proportion in all output working condition type prediction results, namely the real-time working condition type of the train air conditioning unit.
In this embodiment, the preset frequency is 20 groups of data acquired every minute, the preset time is 2.5 minutes, and the environmental parameters and the corresponding internal operation parameters of 50 time nodes are acquired in real time, that is, the real-time parameters of 50 groups of unknown working conditions are acquired, and can be regarded as an environmental parameter time sequence and an internal operation parameter time sequence;
then, for each group of real-time parameters of unknown working conditions, obtaining corresponding residual errors according to the same method in the step 3, wherein all 50 residual errors form a residual error time sequence;
inputting the residual time series (namely 50 residuals) into a fault prediction model of the train air conditioning unit to obtain 50 prediction results related to the working condition types;
and finally, taking one of the 50 prediction results related to the working condition types with the largest ratio as a real-time working condition type prediction result of the train air conditioning unit.
The above embodiments are preferred embodiments of the present application, and those skilled in the art can make various changes or modifications without departing from the general concept of the present application, and such changes or modifications should fall within the scope of the claims of the present application.
Claims (7)
1. An on-the-way identification and diagnosis method for an intelligent train air conditioning unit based on a physical characteristic residual error is characterized by comprising the following steps of:
step 1, obtaining internal operation parameters of a train air conditioning unit respectively operating under various preset environmental parameters and according to normal working conditions; taking the internal operation parameters under each different environment parameter and the corresponding environment parameters as a group of normal parameters; correspondingly acquiring multiple groups of normal parameters for each environment parameter, wherein all the normal parameters form a normal parameter database;
step 2, obtaining internal operation parameters of the train air conditioning unit when the train air conditioning unit operates under various preset environmental parameters and different fault working conditions; taking internal operation parameters of each different environment parameter and different fault working conditions in operation together with corresponding environment parameters as a group of abnormal parameters; the same environmental parameters and the same fault working conditions correspondingly acquire a plurality of groups of abnormal parameters, and all the abnormal parameters form an abnormal parameter database;
step 3, traversing each group of abnormal parameters and normal parameters, finding normal parameters matched with the environmental parameters in a normal parameter database according to the principle of minimum similarity, and performing difference calculation on internal operation parameters in the normal parameters matched with the corresponding environmental parameters of each group of abnormal parameters and normal parameters respectively to obtain corresponding residual errors;
residual errors corresponding to each group of abnormal parameters and normal parameters are used as a training sample;
step 4, training an extreme learning machine model by taking all training samples as input and taking corresponding working condition types as output to obtain a fault prediction model of the train air conditioning unit;
step 5, acquiring an environmental parameter time sequence of the train air conditioning unit within a preset duration and a corresponding internal operation parameter time sequence in real time at a preset frequency, and acquiring a corresponding residual error time sequence according to the step 3; and then inputting the residual error time sequence into a fault prediction model of the train air conditioning unit, and taking one working condition type with the largest proportion in all output working condition type prediction results, namely the real-time working condition type of the train air conditioning unit.
2. The method according to claim 1, wherein the finding of the normal parameters matching the environmental parameters in the normal parameter database according to the principle of minimum similarity includes:
for each group of abnormal parameters to be matched: calculating Euclidean distances between each group of normal parameters in the normal parameter database and the current abnormal parameters to be matched, and taking a group of normal parameters with the minimum Euclidean distance as a matching case of the current abnormal parameters to be matched, namely a group of normal parameters matched with environmental parameters in the current abnormal parameters to be matched;
for each set of normal parameters to be matched: and calculating Euclidean distances between other groups of normal parameters except the normal parameters and the current normal parameters to be matched in the normal parameter database, and taking a group of normal parameters with the minimum Euclidean distance as a matching case of the current normal parameters to be matched, namely a group of normal parameters matched with the environmental parameters in the current normal parameters to be matched.
3. The method of claim 1, wherein training the type of operating condition of the extreme learning machine model comprises: normal operating mode, refrigerant leak, evaporating dish jam, condenser scale deposit, have non-condensable gas and compressor to shut down, other operating mode types except normal operating mode all belong to the fault condition.
4. The method of claim 1, wherein the predetermined frequency in step 5 is 20 sets of data acquired per minute, and the predetermined time is 2.5 minutes.
5. The method of claim 1, wherein the environmental parameters comprise: setting temperature of the air conditioning unit, inlet air dry bulb temperature of the air conditioning unit, inlet air relative humidity of the air conditioning unit and outside air dry bulb temperature; the internal operating parameters include: the air suction temperature of the compressor, the air exhaust temperature of the compressor, the air suction pressure of the compressor, the exhaust pressure of the compressor and the refrigerating capacity of the air conditioner of the train.
6. The method of claim 5, wherein the Euclidean distance is calculated by:
wherein the content of the first and second substances,
tisetting temperature, air conditioning unit inlet air dry bulb temperature, air conditioning unit inlet air relative humidity and outside air dry bulb temperature for the ith group of normal parameters in the normal parameter database respectively; KT (karat)set、KTin、KWinAnd T is respectively set temperature of the air conditioning unit, inlet air dry bulb temperature of the air conditioning unit, inlet air relative humidity of the air conditioning unit and outside air dry bulb temperature in the abnormal parameters to be matched, or is respectively set temperature of the air conditioning unit, inlet air dry bulb temperature of the air conditioning unit, inlet air relative humidity of the air conditioning unit and outside air dry bulb temperature in the normal parameters to be matched.
7. The method of claim 1, wherein the parameters of the extreme learning machine optimized by the differential evolution algorithm include input weight vectors and hidden layer neuron bias vectors, and the optimization process is as follows:
step 4.1, parameter setting and population initialization: setting maximum evolution algebra GmaxPopulation size NP, scaling factor F and crossover probability Pcr(ii) a Representing each individual in the population as a target vector r ═ W b consisting of the input weight vector and the hidden layer neuron bias vector]For the input weight vector of the parameter vector, each element is in the range [0, 1 ]]Performing random initialization on each element in hidden layer neuron bias vector of parameter vector in range of [ -5, 5]Carrying out random initialization; taking the variance between the predicted value and the true value of the extreme learning machine as a fitness function;
step 4.2, making the evolution algebra G equal to 0; the target vector r is initialized to p (0);
step 4.3, making the evolution algebra G equal to G + 1; p (G +1) ═ p (G);
step 4.4, randomly selecting 3 individuals r1, r2 and r3 in the population outside the target vector p (G);
step 4.5, carrying out differential variation operation to generate a variation vector;
r*(G)＝r1(G)+F*(r2(G)-r3(G))
wherein F is called a scaling factor and is a constant; r is*Is a variation vector;
step 4.6, performing cross operation on the target vector and the variation vector to generate a test vector;
specifically, each component of the test vector is subjected to a crossover operation:
wherein the content of the first and second substances,
is the G-th generation test vector component; pcrIs the cross probability, is a constant;
is the G generation variation vector component;
is a component of the G-th generation target vector.
4.7, calculating the fitness value of the test vector, and performing comparison selection operation;
respectively substituting the test vector and the target vector into an extreme learning machine to calculate corresponding fitness function values, and taking the fitness function value which is more optimal as a new target vector;
wherein, r (G) is the G-th generation target vector; and f is a fitness function.
Step 4.8, if G ═ GmaxAnd ending the optimization process and outputting the network weight W of the current target vector and the hidden layer neuron offset b, otherwise, turning to the step 4.3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010558477.7A CN111546854B (en) | 2020-06-18 | 2020-06-18 | On-road identification and diagnosis method for intelligent train air conditioning unit |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010558477.7A CN111546854B (en) | 2020-06-18 | 2020-06-18 | On-road identification and diagnosis method for intelligent train air conditioning unit |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111546854A CN111546854A (en) | 2020-08-18 |
| CN111546854B true CN111546854B (en) | 2021-08-24 |
Family
ID=72007054
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010558477.7A Active CN111546854B (en) | 2020-06-18 | 2020-06-18 | On-road identification and diagnosis method for intelligent train air conditioning unit |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111546854B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112696791A (en) * | 2020-12-25 | 2021-04-23 | 山东朗进科技股份有限公司 | Rail vehicle air conditioning unit refrigerant leakage early warning method |
| CN112907781B (en) * | 2021-02-07 | 2022-08-19 | 中国人民解放军国防科技大学 | System fault diagnosis method and device, computer equipment and storage medium |
| CN113837479B (en) * | 2021-09-28 | 2024-03-15 | 江苏湛德医疗用品有限公司 | Early warning method and system for monitoring running state of target equipment |
| CN114312213B (en) * | 2021-12-14 | 2023-10-17 | 华人运通(江苏)技术有限公司 | Method, device, equipment and medium for predicting abnormal refrigerant capacity of vehicle |
| CN115129687A (en) * | 2022-03-16 | 2022-09-30 | 希望知舟技术(深圳)有限公司 | Abnormal condition database management-based method, related device, medium and program |
| CN115563622B (en) * | 2022-09-29 | 2024-03-12 | 国网山西省电力公司 | Method, device and system for detecting operation environment |
| CN116226239B (en) * | 2023-05-06 | 2023-07-07 | 成都瑞雪丰泰精密电子股份有限公司 | Data-driven-based state monitoring method for spindle system of machining center |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105046210A (en) * | 2015-07-01 | 2015-11-11 | 西安理工大学 | Multi-label diagnostic method for train air conditioning faults |
| CN105135591A (en) * | 2015-07-01 | 2015-12-09 | 西安理工大学 | Train air conditioning unit fault diagnosing method based on multi-classification strategy |
| CN108507117A (en) * | 2017-10-13 | 2018-09-07 | 上海智容睿盛智能科技有限公司 | A kind of Air-conditioning system sensor method for diagnosing faults based on wavelet neural network |
| CN110173872A (en) * | 2019-04-22 | 2019-08-27 | 广东美博制冷设备有限公司 | The misoperation detection method and device of air conditioner |
| US10474932B2 (en) * | 2016-09-01 | 2019-11-12 | Uptake Technologies, Inc. | Detection of anomalies in multivariate data |
-
2020
- 2020-06-18 CN CN202010558477.7A patent/CN111546854B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105046210A (en) * | 2015-07-01 | 2015-11-11 | 西安理工大学 | Multi-label diagnostic method for train air conditioning faults |
| CN105135591A (en) * | 2015-07-01 | 2015-12-09 | 西安理工大学 | Train air conditioning unit fault diagnosing method based on multi-classification strategy |
| US10474932B2 (en) * | 2016-09-01 | 2019-11-12 | Uptake Technologies, Inc. | Detection of anomalies in multivariate data |
| CN108507117A (en) * | 2017-10-13 | 2018-09-07 | 上海智容睿盛智能科技有限公司 | A kind of Air-conditioning system sensor method for diagnosing faults based on wavelet neural network |
| CN110173872A (en) * | 2019-04-22 | 2019-08-27 | 广东美博制冷设备有限公司 | The misoperation detection method and device of air conditioner |
Non-Patent Citations (1)
| Title |
|---|
| 列车空调机组故障检修中SVM的应用;罗浩;《中国优秀硕士学位论文全文数据库 信息科技II辑》;20100515(第5期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111546854A (en) | 2020-08-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111546854B (en) | On-road identification and diagnosis method for intelligent train air conditioning unit | |
| US20200150622A1 (en) | Method for detecting abnormity in unsupervised industrial system based on deep transfer learning | |
| CN108931387B (en) | Fault diagnosis method based on multi-sensor signal analysis and capable of providing accurate diagnosis decision | |
| Mulumba et al. | Robust model-based fault diagnosis for air handling units | |
| US20200184131A1 (en) | A method for prediction of key performance parameter of an aero-engine transition state acceleration process based on space reconstruction | |
| CN111723925B (en) | Fault diagnosis method, device, equipment and medium for on-road intelligent train air conditioning unit | |
| Wang et al. | Feature selection based on Bayesian network for chiller fault diagnosis from the perspective of field applications | |
| CN108334059A (en) | Based on particle cluster algorithm Optimized BP Neural Network Symbolic fault diagnosis method | |
| CN109213127A (en) | A kind of HVAC system gradual failure diagnostic method based on deep learning | |
| Zhou et al. | Data-driven fault diagnosis for residential variable refrigerant flow system on imbalanced data environments | |
| CN111351668A (en) | Diesel engine fault diagnosis method based on optimized particle swarm algorithm and neural network | |
| CN114484731A (en) | Method and device for diagnosing faults of central air conditioner based on stacking fusion algorithm | |
| Li et al. | Diagnosis for multiple faults of chiller using ELM-KNN model enhanced by multi-label learning and specific feature combinations | |
| CN112462001B (en) | Gas sensor array model calibration method for data amplification based on condition generation countermeasure network | |
| CN112766047B (en) | Fault diagnosis method for refrigeration system and refrigeration device | |
| CN110715405A (en) | Air conditioner refrigerating capacity detection method based on BP neural network fitting model | |
| Li et al. | Machine learning based diagnosis strategy for refrigerant charge amount malfunction of variable refrigerant flow system | |
| Zhang et al. | Causal discovery and inference-based fault detection and diagnosis method for heating, ventilation and air conditioning systems | |
| CN113822359B (en) | Heat pump system self-adaptive fault diagnosis method based on residual data scaling strategy | |
| CN115618506A (en) | Method for predicting power of single-shaft combined cycle gas turbine | |
| CN113688895A (en) | Method and system for detecting abnormal firing zone of ceramic roller kiln based on simplified KECA | |
| CN113051530A (en) | KDE-FA-based cold water unit fault feature characterization method | |
| CN113465930B (en) | Gas turbine multi-sensor fault detection method based on hybrid method | |
| CN111967489B (en) | Manufacturing process abnormity monitoring method based on quality data manifold characteristics | |
| Liu et al. | Feature Selection for Chiller Fault Detection and Diagnosis Based on Grey Similitude Degree of Incidence. |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |