CN111898725A - Air conditioning system sensor fault detection method and device and electronic equipment - Google Patents
Air conditioning system sensor fault detection method and device and electronic equipment Download PDFInfo
- Publication number
- CN111898725A CN111898725A CN202010647428.0A CN202010647428A CN111898725A CN 111898725 A CN111898725 A CN 111898725A CN 202010647428 A CN202010647428 A CN 202010647428A CN 111898725 A CN111898725 A CN 111898725A
- Authority
- CN
- China
- Prior art keywords
- conditioning system
- particle
- fault detection
- kernel function
- air conditioning
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 72
- 238000004378 air conditioning Methods 0.000 title claims abstract description 69
- 239000002245 particle Substances 0.000 claims abstract description 88
- 238000005457 optimization Methods 0.000 claims abstract description 44
- 238000000034 method Methods 0.000 claims abstract description 30
- 238000011156 evaluation Methods 0.000 claims abstract description 12
- 238000013178 mathematical model Methods 0.000 claims abstract description 11
- 239000011159 matrix material Substances 0.000 claims abstract description 11
- 238000000513 principal component analysis Methods 0.000 claims abstract description 5
- 230000006870 function Effects 0.000 claims description 85
- 238000012549 training Methods 0.000 claims description 9
- 238000012360 testing method Methods 0.000 claims description 5
- 238000005259 measurement Methods 0.000 claims description 3
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 3
- 230000006399 behavior Effects 0.000 claims description 2
- 230000005540 biological transmission Effects 0.000 claims description 2
- 230000008901 benefit Effects 0.000 abstract description 7
- 238000005265 energy consumption Methods 0.000 description 9
- 230000006872 improvement Effects 0.000 description 8
- 238000003745 diagnosis Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 6
- 238000010438 heat treatment Methods 0.000 description 4
- NUFBIAUZAMHTSP-UHFFFAOYSA-N 3-(n-morpholino)-2-hydroxypropanesulfonic acid Chemical compound OS(=O)(=O)CC(O)CN1CCOCC1 NUFBIAUZAMHTSP-UHFFFAOYSA-N 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 238000012847 principal component analysis method Methods 0.000 description 3
- 238000005034 decoration Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 230000008014 freezing Effects 0.000 description 2
- 238000007710 freezing Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004134 energy conservation Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002431 foraging effect Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 238000009423 ventilation Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D18/00—Testing or calibrating apparatus or arrangements provided for in groups G01D1/00 - G01D15/00
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Air Conditioning Control Device (AREA)
Abstract
The invention discloses a method and a device for detecting faults of a sensor of an air conditioning system and electronic equipment, wherein the method comprises the following steps: determining a system evaluation index and establishing a target function so as to determine a fitness function; selecting a Gaussian radial basis kernel function and a polynomial kernel function structure to establish a hybrid kernel function matrix for detecting the faults of the sensors of the air-conditioning system, and establishing a hybrid kernel function-based KPCA (kernel principal component analysis) mathematical model for describing the running state of the air-conditioning system and detecting the faults of the sensors; and optimizing the parameters of the mixed kernel function by utilizing a multi-objective particle swarm algorithm to obtain the fault detection optimization of the sensor of the air conditioning system. The method introduces the multi-target particle swarm algorithm to solve the problem of the fault detection optimization of the air conditioning system sensor, improves the hybrid algorithm to solve the target problem, has certain advantages, and can realize the fault detection optimization of the air conditioning system sensor.
Description
Technical Field
The invention belongs to the field of air conditioning system fault detection, relates to a fault detection optimization method, and particularly relates to a method and a device for detecting faults of a sensor of an air conditioning system and electronic equipment.
Background
Heating, Ventilating and Air Conditioning (HVAC) systems are used as important components in intelligent building systems, and have been widely applied and rapidly and remarkably developed along with the development of intelligent buildings. Nowadays, higher requirements are provided for the intelligent integration of indoor environment quality and the energy consumption of buildings in intelligent buildings, and then the requirements for comfort, intelligence and energy conservation are continuously improved on the premise of stable operation of an HVAC system, so that the HVAC system is more and more complicated. In an increasingly complex heating, ventilating and air conditioning system, the probability of various faults is gradually increased along with the lapse of time due to self defects, overload operation, lack of operation and maintenance and the like. And the existence of the fault not only causes energy waste but also influences the indoor environment quality, and even shortens the service life of the equipment. In view of various considerations, a Fault Detection and Diagnosis (FDD) system plays a crucial role in timely and accurate Fault Detection and identification, and becomes an essential component of an HVAC system. The HVAC system fault diagnosis is to identify and judge fault positions (areas), fault types and fault reasons by matching HVAC existing knowledge with an FDD method, wherein a fault component is determined to be a core problem. The application of the fault diagnosis technology in the system not only greatly helps the equipment to be overhauled and maintained; meanwhile, theoretical analysis and actual data show that the building energy consumption accounts for about 30% of the total social energy consumption, the HVAC system in the building is the most important energy consumption equipment, the operation energy consumption can account for 50% -60% of the building energy consumption, and the trend of the increase year by year is shown. Meanwhile, relevant data show that when the HVAC system breaks down, the energy consumption of the system can be increased by 15% -30%; by implementing fault detection and diagnosis to optimize the operation of the heating, ventilating and air conditioning system, the building energy consumption can be reduced by 20-30%, and the air conditioning system energy consumption can be reduced by 10-40%.
Disclosure of Invention
In order to solve the problem that parameter selection is difficult to achieve the optimum in the implementation process of an algorithm in the prior art, the invention aims to provide a method, a device and electronic equipment for detecting the fault of a sensor of an air conditioning system.
In order to realize the task, the invention adopts the following technical solution:
a method for detecting faults of sensors of an air conditioning system comprises the following steps:
determining a system evaluation index and establishing a target function so as to determine a fitness function;
selecting a Gaussian radial basis kernel function and a polynomial kernel function structure to establish a hybrid kernel function matrix for detecting the faults of the sensors of the air-conditioning system, and establishing a hybrid kernel function-based KPCA (kernel principal component analysis) mathematical model for describing the running state of the air-conditioning system and detecting the faults of the sensors;
and optimizing the parameters of the mixed kernel function by utilizing a multi-objective particle swarm algorithm to obtain the fault detection optimization of the sensor of the air conditioning system.
As a further improvement of the invention, according to the target index of the air conditioner, determining the fault detection evaluation index of the sensor of the air conditioning system; the target indexes comprise root mean square error, fault detection rate and principal component number.
As a further improvement of the present invention, the target function corresponding to the target index is as follows:
root mean square error
In the formula (1), rho is an influence coefficient of a mixed kernel function, d is an order parameter of a polynomial, and sigma is a kernel width parameter of a Gaussian radial basis; y isactualActual value of ith sample, yforecast-a predicted value of the ith sample,
-the SVM prediction function obtained under given parameters and training samples, n-the number of training samples;
fault detection rate
In the formulae (2), (3) and (4), n1—T2The number of detected faults under the statistic, n is the total number of set faults,
-the eigenvectors corresponding to the eigenvalues of the kernel matrix,
-kernel function, tk—T2Statistics, p-principal component number, N-sample number; formula (3) is a judgment formula, if n1>1, then represents the ith T2The statistic value is greater than T2Counting the control limit, namely increasing the number of detected faults by 1, and so on; x is training sample data, and y is test sample data; rho is the influence coefficient of the mixed kernel function, d is the order parameter of the polynomial, and sigma is the kernel width parameter of the Gaussian radial basis.
Principal component number III
And (5) the ratio of the sum of the first characteristic values to the sum of the total characteristic values exceeds E, when the ratio is greater than or equal to the set E value, the number of the principal elements is increased by 1, and the E is 85 percent.
As a further improvement of the present invention, the mixing kernel function is as follows:
in the formula (6), x is training sample data, and y is test sample data; rho is the influence coefficient of the mixing kernel function; d is the order parameter of the polynomial and σ is the kernel width parameter of the Gaussian radial basis.
As a further improvement of the invention, the KPCA mathematical model based on the mixed kernel function is as follows:
X＝[Tfre,Tsup,Trtn,Hfre,Hsup,Hrtn,cw]T(7)
in the formula (7), a state model is established mainly by adopting the measurement value of a typical temperature and humidity sensor in the air conditioning system; t isfreFresh air temperature, TsupSupply air temperature, TrtnReturn air temperature HfreFresh air humidity, HsupHumidity of air supply HrtnReturn air humidity and cw freezing water valve opening.
As a further improvement of the present invention, the optimizing the mixed kernel function parameters by using the multi-objective particle swarm optimization specifically includes:
determining the optimal position under the current searching progress, the optimal position of the current position and the optimal position of the whole population of each particle according to the fitness determined by the objective function; performing iterative optimization according to the relative position relationship between the current position and the optimal position as well as the optimal position of the whole population;
comparing the determined fitness with the particle individual extreme value, and judging whether the current position is used as the optimal position; comparing the determined fitness with the particle global extreme value, and judging whether to update the optimal position of the whole population;
and judging whether the termination condition is met, if so, outputting an optimal result, and if not, continuously and repeatedly generating a new solution to replace the old solution.
As a further improvement of the present invention, the optimization means: adopting continuous value coding, setting each parameter and determining the population number; initializing and setting relevant problems involved in the particle optimization process, searching for a sufficient solution space when the number of particles is 5-10 times of the dimension of the particle, and obtaining the maximum velocity vmaxRefers to the maximum velocity set by the particle when performing the target search; v. ofmaxThe setting of the size determines the searching capability of the particles, and has great influence on the optimizing result; maximum velocity vmaxObtained by the following formula:
vmax＝xmax-xmin(8)
in the formula (8), vmaxMaximum speed, x, set by the particles for the target searchmaxActual value, x, at the time of searching for a target at the maximum speed of the particlemin-actual values when the particle minimum velocity searches for targets.
As a further improvement of the present invention, the multi-target particle swarm algorithm is that each particle performs information transmission in a certain space according to a certain rule or mode, and continuously performs self-organization action for self-correction according to the deviation between the actual position and the optimal position, and the following formula is adopted to operate the particles:
vid＝ω*vid+c1*r1(pid-xid)+c2*r2(pgd-xgd) (9)
xid＝xid+vid(10)
in formulas (9) and (10), i is the total number of particles in the population; v. ofid-iterating the d-dimensional component of the velocity vector of the particle i; x is the number ofid-iterating the d-dimensional component of the position in the particle i search space; p is a radical ofid—pbestThe d-th dimension component of (1); p is a radical ofgd—gbestThe d-th dimension component of (1); c. C1、c2-a learning factor; r is1、r2-a random function; ω — inertial weight.
As a further improvement of the invention, the particle fitness is determined by first determining the particle XiIndividual extreme value p ofid，
Initial global extremum pgdI.e. p corresponding to the particle with the best fitness valueid(ii) a According to V of the particleiAnd XiThe fitness value is continuously updated, so as to update the global extremum pgd；
Determining an individual optimum extremum pidAnd the current fitness value ppresent: if p ispresent>pidThen p isid＝ppresent(ii) a Determining a global optimal position pgd: if p ispresent>pgdThen p isgd＝ppresent；
If the end condition is met, optimizing, ending and outputting the result; if not, the method shifts to the particle fitness determination.
An air conditioning system sensor fault detection system comprising:
the target function establishing module is used for determining a system evaluation index and establishing a target function so as to determine a fitness function;
the mixed kernel function establishing module is used for selecting a Gaussian radial basis kernel function and a polynomial kernel function structure to establish a sensor fault detection mixed kernel function matrix of the air-conditioning system and establishing a KPCA (kernel-based interference cancellation) mathematical model which describes the running state of the air-conditioning system and the sensor fault detection and is based on the mixed kernel function;
and the detection optimization module optimizes the mixed kernel function parameters by using a multi-objective particle swarm algorithm to obtain the fault detection optimization of the air conditioning system sensor.
An electronic device, comprising:
one or more processors;
a memory;
one or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, the one or more applications configured to: and executing the steps of the air conditioning system sensor fault detection method.
A computer readable storage medium for storing computer instructions which, when executed on a computer, cause the computer to perform the steps of the air conditioning system sensor fault detection method described above.
Compared with the prior art, the invention has the following advantages:
according to the air conditioning system sensor fault detection method, three target indexes including the root mean square error, the fault detection rate and the principal component number are considered comprehensively, optimization of a target function can be achieved through a multi-target particle swarm algorithm, and the detection rate of the air conditioning system sensor fault based on a KPCA method is further improved. And the basic KPCA method is improved, the parameters of a mixed kernel function in the KPCA are further optimized, the KPCA algorithm can better play the role, and the method has superiority in further optimizing the fault detection of the sensor in the air conditioning system and improving the fault rate.
Drawings
FIG. 1 is a schematic diagram of a mixed kernel function parameter optimization process of the MOPSO algorithm of the present invention;
FIG. 2 is a root mean square error optimization graph of an embodiment of the present invention;
FIG. 3 is a graph of principal component optimization for an embodiment of the present invention;
FIG. 4 shows an embodiment of the present invention based on T2Optimizing a curve graph of the statistic detection rate;
FIG. 5 is a schematic diagram of an air conditioning system sensor fault detection system according to the present invention.
Detailed Description
The invention is described in further detail below with reference to the figures and the detailed description.
Referring to fig. 1, a schematic diagram of a mixed kernel function parameter Optimization process of the MOPSO algorithm of the present invention, a Particle Swarm Optimization (PSO) is a Swarm Optimization algorithm, which has a special advantage for target Optimization, especially multi-target Optimization, and is proposed by Eberhart and Kennedy in 1995. The PSO algorithm realizes the process of iteratively searching the optimal solution from random solution by simulating foraging of the bird group.
The invention aims to overcome the defects of the prior art and provides a fault detection method for a sensor of an air conditioning system, namely the fault detection method for the air conditioning system based on an MOPSO-KPCA algorithm. The multi-target particle swarm algorithm has the advantages of high search precision and high convergence speed when solving the nonlinear optimization problem, and the multi-target particle swarm algorithm is combined with a kernel principal component analysis method and used for detecting the faults of the sensor of the air conditioning system, so that the detection time can be shortened, the fault detection rate of the sensor is further improved, and the optimization of the fault detection effect of the sensor is realized. The technical scheme of the invention is as follows:
s100, comprehensively considering three target indexes including a root mean square error, a fault detection rate and a principal component number, and determining a fault detection evaluation index of a sensor of the air conditioning system;
s200, performing system modeling, selecting a Gaussian radial basis kernel function and a polynomial kernel function to construct a mixed kernel function, and establishing a KPCA (kernel principal component analysis) mathematical model based on the mixed kernel function by using a kernel function matrix;
s300, establishing a fault detection state model of the air conditioning system sensor based on the hybrid kernel function, and optimizing parameters of the hybrid kernel function by utilizing a multi-target particle swarm algorithm;
s400, initializing parameters including the size of the initialized population and the initial velocity V of particlesiCurrent position X of particlei；
S500, determining the optimal position p under the current search progress according to the fitness determined by the objective functionbest(ii) a In determining the optimum position pbestAnd the current position XiMeanwhile, each particle can obtain the optimal position g of the whole populationbest；
S600, then determining XiAnd pbest、gbestThe relative position relation comprises information such as distance, direction and the like, and finally, the next flight route is realized according to the feedback information, namely iterative optimization is carried out;
s700, comparing the determined fitness with the particle individual extreme value, and judging whether X can be used or notiAs pbest(ii) a Comparing the determined fitness with the particle global extreme value to judge whether the g can be updated or notbest；
And S800, finally judging whether a termination condition is met, outputting an optimal result if the termination condition is met, and continuously and repeatedly generating a new solution to replace the old solution and the subsequent steps if the termination condition is not met.
Each step is explained in detail below.
When the multi-target particle swarm algorithm is combined with a core principal component analysis method to be used for the fault detection problem of the sensor of the air conditioning system, the system state and the diagnosis strategy when the core principal component analysis is used for fault detection are fully considered, namely model constraint conditions:
in the fault diagnosis of the heating ventilation air conditioning system sensor by applying the kernel principal component analysis method, the following constraint conditions need to be considered according to a fault diagnosis model:
(1) when the air conditioning system detects faults of the sensor, the current system running state and the working condition are considered;
(2) the value of the fault detection rate is [0,1], and the larger the fault detection rate value is, the better the fault detection rate value is; the fault detection rate value range further standardizes the value range of the optimization target;
(3) the number of the principal elements is determined by the contribution rate of the principal elements, the value of the number of the principal elements is a positive integer larger than 0, and the smaller the number of the principal elements is, the better the number of the principal elements is under the condition of ensuring the integrity of information;
(4) for the parameters in the mixed kernel function, the value range is rho epsilon [0,0.9], d is a positive integer larger than 0, and sigma is larger than 0.
Step 1: the invention firstly determines the fault detection evaluation index of the air conditioning system, comprehensively considers three target indexes of root mean square error, fault detection rate and principal component number, and determines the fault detection evaluation index of the air conditioning system sensor, wherein the target functions corresponding to the target indexes are as follows:
root Mean Square Error (RMSE)
In the formula (1), rho is an influence coefficient of a mixed kernel function, d is an order parameter of a polynomial, and sigma is a kernel width parameter of a Gaussian radial basis; y isactualActual value of ith sample, yforecast-a predicted value of the ith sample,
-the SVM prediction function obtained under given parameters and training samples, n-the number of training samples;
fault Detection Rate (FDR)
In the formulae (2), (3) and (4), n1—T2The number of detected faults under the statistic, n is the total number of set faults,
-the eigenvectors corresponding to the eigenvalues of the kernel matrix,
-kernel function, tk—T2Statistics, p-principal component number, N-sample number; formula (3) is a judgment formula, if n1>1, then represents the ith T2The statistic value is greater than T2Counting the control limit, namely increasing the number of detected faults by 1, and so on; x is training sample data, and y is test sample data; rho is the influence coefficient of the mixed kernel function, d is the order parameter of the polynomial, and sigma is the kernel width parameter of the Gaussian radial basis.
Principal component Number (NPC)
The expression (5) shows that the ratio of the sum of the first characteristic values to the sum of the total characteristic values exceeds E, when the ratio is larger than or equal to the set E value, the number of the principal elements is increased by 1, and E is usually 85%.
Step 2: determining a kernel matrix mathematical model in KPCA based on a mixed kernel function, wherein the mixed kernel function is as follows:
in the formula (6), x is training sample data, and y is test sample data; rho is an influence coefficient of a mixed kernel function, and adaptability of system parameter drift is enhanced. d is the order parameter of the polynomial and σ is the kernel width parameter of the Gaussian radial basis. For the mixed kernel function, the above parameters have a crucial influence on the modeling. Therefore, in order to obtain a better diagnostic model, it is necessary to optimize the parameters ρ, d, and σ in the mixed kernel function according to the actual operating data of the system.
And step 3: determining a hybrid kernel function based air conditioning system sensor fault detection state model as follows:
X＝[Tfre,Tsup,Trtn,Hfre,Hsup,Hrtn,cw]T(7)
in the formula (7), a state model is established mainly by using the measurement values of typical temperature and humidity sensors in the air conditioning system. T isfreFresh air temperature, TsupSupply air temperature, TrtbReturn air temperature HfreFresh air humidity, HsupHumidity of air supply HrtnReturn air humidity and cw freezing water valve opening.
And 4, step 4: adopting continuous value coding, setting each parameter and determining the population number; initializing relevant problems involved in particle optimization, e.g. V for particle searchiAnd vmaxLearning factor weight ratio and range, maximum iteration number; in general, when the number of particles is 5-10 times of the dimension of the particle, enough solution space can be searched, and the maximum velocity vmaxRefers to the maximum velocity set by the particle when performing the target search; v. ofmaxThe set size determines the searching capability of the particles, and has a large influence on the optimizing result. Maximum velocity vmaxCan be obtained by the following formula:
vmax＝xmax-xmin(8)
in the formula (8), vmaxMaximum speed, x, set by the particles for the target searchmaxActual value, x, at the time of searching for a target at the maximum speed of the particlemin-actual values when the particle minimum velocity searches for targets.
Step 6: the MOPSO algorithm can be regarded as a self-organizing behavior that each particle carries out information transfer in a certain space according to a certain rule or mode and corrects itself continuously according to the deviation of an actual position and an optimal position. The following formula can be used to operate on the particles:
vid＝ω*vid+c1*r1(pid-xid)+c2*r2(pgd-xgd) (9)
xid＝xid+vid(10)
in formulas (9) and (10), i is the total number of particles in the population; v. ofid-iterating the d-dimensional component of the velocity vector of the particle i; xid-d-dimension component of position in iterative particle i search space; p is a radical ofid—pbestThe d-th dimension component of (1); p is a radical ofgd—gbestThe d-th dimension component of (1); c. C1、c2-a learning factor; r is1、r2-a random function; ω — inertial weight.
And 7: determining the particle fitness: first determining XiI.e. individual extrema p of the particlesid. Determining fitness value of the particle, then initial global extremum pgdI.e. p corresponding to the particle with the best fitness valueid(ii) a According to V of the particleiAnd XiThe fitness value is continuously updated, so as to update the global extremum pgd；
And 8: determining an individual optimum extremum pidAnd the current fitness value ppresent: if p ispresent>pidThen p isid＝ppresent(ii) a Determining a global optimal position pgd: if p ispresent>pgdThen p isgd＝ppresent；
The optimized root mean square error optimization curve is shown in FIG. 2, the optimized principal component number optimization curve is shown in FIG. 3, and the optimized root mean square error is based on T2The statistics detection rate optimization graph is shown in fig. 4, and the non-optimized fault detection rate is shown in table 1; the improved algorithm has the advantage over the basic algorithm in solving the problem of fault detection optimization of the air conditioning system sensor.
TABLE 1
Simulation results show that the multi-target particle swarm algorithm is introduced to solve the problem of the fault detection optimization of the air conditioning system sensor, and the hybrid algorithm is improved to solve the target problem, so that the method has certain advantages and can realize the fault detection optimization of the air conditioning system sensor.
In summary, the method for detecting the fault of the sensor of the air conditioning system of the present invention includes determining a system evaluation index and establishing a target function to determine a fitness function; and establishing a sensor fault detection kernel function matrix of the air conditioning system, and describing the operating state of the air conditioning system and a mathematical model for detecting the sensor fault. Setting parameters in a particle swarm algorithm and determining the population number; initializing relevant problems involved in the particle optimization process, and determining X firstiI.e. individual extrema p of the particlesid. Determining fitness value of the particle, then initial global extremum pgdI.e. p corresponding to the particle with the best fitness valueid(ii) a According to V of the particleiAnd XiThe fitness value is continuously updated, so as to update the global extremum pgd(ii) a If p ispresent>pidThen p isid＝ppresent(ii) a If p ispresent>pgdThen p isgd＝ppresent(ii) a If the end condition is met, optimizing, ending and outputting the result; if not, the method shifts to the particle fitness determination. The method introduces the multi-target particle swarm algorithm to solve the problem of the fault detection optimization of the air conditioning system sensor, improves the hybrid algorithm to solve the target problem, has certain advantages, and can realize the fault detection optimization of the air conditioning system sensor.
Further, as shown in fig. 5, an embodiment of the present invention further provides an air conditioning system sensor fault detection system, including:
the target function establishing module is used for determining a system evaluation index and establishing a target function so as to determine a fitness function;
the mixed kernel function establishing module is used for selecting a Gaussian radial basis kernel function and a polynomial kernel function structure to establish a sensor fault detection mixed kernel function matrix of the air-conditioning system and establishing a KPCA (kernel-based interference cancellation) mathematical model which describes the running state of the air-conditioning system and the sensor fault detection and is based on the mixed kernel function;
and the detection optimization module optimizes the mixed kernel function parameters by using a multi-objective particle swarm algorithm to obtain the fault detection optimization of the air conditioning system sensor.
The detection optimization module is specifically configured to:
determining the optimal position under the current searching progress, the optimal position of the current position and the optimal position of the whole population of each particle according to the fitness determined by the objective function; performing iterative optimization according to the relative position relationship between the current position and the optimal position as well as the optimal position of the whole population;
comparing the determined fitness with the particle individual extreme value, and judging whether the current position is used as the optimal position; comparing the determined fitness with the particle global extreme value, and judging whether to update the optimal position of the whole population;
and judging whether the termination condition is met, if so, outputting an optimal result, and if not, continuously and repeatedly generating a new solution to replace the old solution.
An embodiment of the present application provides an electronic device, which includes:
one or more processors;
a memory;
one or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, the one or more applications configured to: performing the steps of the air conditioning system sensor fault detection method according to any one of claims 1 to 7.
The memory 503 may be, but is not limited to, a ROM or other type of static storage device that can store static information and instructions, a RAM or other type of dynamic storage device that can store information and instructions, an EEPROM, a CD-ROM or other optical disk storage, optical disk storage (including compact disk, laser disk, optical disk, digital versatile disk, blu-ray disk, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer.
Embodiments of the present application also provide a computer-readable storage medium for storing computer instructions, which when executed on a computer, enable the computer to perform the steps of the air conditioning system sensor fault detection method according to any one of the above claims 1 to 7.
It should be understood that, although the steps in the flowcharts of the figures are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and may be performed in other orders unless explicitly stated herein. Moreover, at least a portion of the steps in the flow chart of the figure may include multiple sub-steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, which are not necessarily performed in sequence, but may be performed alternately or alternately with other steps or at least a portion of the sub-steps or stages of other steps.
The foregoing is only a partial embodiment of the present application, and it should be noted that, for those skilled in the art, several modifications and decorations can be made without departing from the principle of the present application, and these modifications and decorations should also be regarded as the protection scope of the present application.
Claims (10)
1. A method for detecting faults of sensors of an air conditioning system is characterized by comprising the following steps:
determining a system evaluation index and establishing a target function so as to determine a fitness function;
selecting a Gaussian radial basis kernel function and a polynomial kernel function structure to establish a hybrid kernel function matrix for detecting the faults of the sensors of the air-conditioning system, and establishing a hybrid kernel function-based KPCA (kernel principal component analysis) mathematical model for describing the running state of the air-conditioning system and detecting the faults of the sensors;
and optimizing the parameters of the mixed kernel function by utilizing a multi-objective particle swarm algorithm to obtain the fault detection optimization of the sensor of the air conditioning system.
2. The method of claim 1, wherein an air conditioning system sensor fault detection evaluation index is determined according to a target index of an air conditioner; the target indexes comprise root mean square error, fault detection rate and principal component number.
3. The method of claim 1, wherein the mixing kernel is as follows:
in the formula (6), x is training sample data, and y is test sample data; p is the influence coefficient of the mixed kernel function, d is the order parameter of the polynomial, and σ is the kernel width parameter of the Gaussian radial basis.
4. The method of claim 1, wherein the hybrid kernel function based KPCA mathematical model is as follows:
X＝[Tfre，Tsup，Trtn，Hfre，Hsup，Hrtn，cw]T(7)
in the formula (7), a state model is established mainly by adopting the measurement value of a typical temperature and humidity sensor in the air conditioning system; t isfreFresh air temperature, Tsup-temperature of supply air, TrtnReturn air temperature, HfreFresh air humidity, HsupSupply air humidity Hrtn-return air humidity and cw-chilled water valve opening.
5. The method of claim 1, wherein the optimizing the mixed kernel function parameters using the multi-objective particle swarm optimization specifically comprises:
determining the optimal position under the current searching progress, the optimal position of the current position and the optimal position of the whole population of each particle according to the fitness determined by the objective function; performing iterative optimization according to the relative position relationship between the current position and the optimal position as well as the optimal position of the whole population;
comparing the determined fitness with the particle individual extreme value, and judging whether the current position is used as the optimal position; comparing the determined fitness with the particle global extreme value, and judging whether to update the optimal position of the whole population;
and judging whether the termination condition is met, if so, outputting an optimal result, and if not, continuously and repeatedly generating a new solution to replace the old solution.
6. The method of claim 5, wherein the multi-objective particle swarm optimization is that each particle performs information transmission in a certain rule or mode in a certain space and operates on the particle by adopting the following formula according to the self-organization behavior of correcting the particle according to the deviation between the actual position and the optimal position:
vid＝ω*vid+c1*r1(pid-xid)+c2*r2(pgd-xgd) (9)
xid＝xid+vid(10)
in formulas (9) and (10), i-the total number of particles in the population; v. ofid-iterating the d-dimensional component of the particle i's airspeed vector; x is the number ofid-iterating the d-dimensional component of the position in the particle i search space; p is a radical ofid-pbestThe d-th dimension component of (1); p is a radical ofgd—gbestThe d-th dimension component of (1); c. C1、c2-a learning factor; r is1、r2-a random function; ω -inertial weight.
7. The method of claim 5, wherein the particle fitness is a first determination of particle XiIndividual extreme value p ofid，
Initial global extremum pgdI.e. p corresponding to the particle with the best fitness valueid(ii) a According to V of the particleiAnd XiThe fitness value is continuously updated, thereby updating the wholeLocal extremum pgd；
Determining an individual optimum extremum pidAnd the current fitness value ppresent: if p ispresent＞pidThen p isid＝ppresent(ii) a Determining a global optimal position pgd: if p ispresent＞pgdThen p isgd＝ppresent；
If the end condition is met, optimizing, ending and outputting the result; if not, the method shifts to the particle fitness determination.
8. An air conditioning system sensor fault detection system, comprising:
the target function establishing module is used for determining a system evaluation index and establishing a target function so as to determine a fitness function;
the mixed kernel function establishing module is used for selecting a Gaussian radial basis kernel function and a polynomial kernel function structure to establish a sensor fault detection mixed kernel function matrix of the air-conditioning system and establishing a KPCA (kernel-based interference cancellation) mathematical model which describes the running state of the air-conditioning system and the sensor fault detection and is based on the mixed kernel function;
and the detection optimization module optimizes the mixed kernel function parameters by using a multi-objective particle swarm algorithm to obtain the fault detection optimization of the air conditioning system sensor.
9. An electronic device, comprising:
one or more processors;
a memory;
one or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, the one or more applications configured to: performing the steps of the air conditioning system sensor fault detection method according to any one of claims 1 to 7.
10. A computer readable storage medium for storing computer instructions which, when executed on a computer, cause the computer to perform the steps of the air conditioning system sensor fault detection method of any of the preceding claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010647428.0A CN111898725A (en) | 2020-07-07 | 2020-07-07 | Air conditioning system sensor fault detection method and device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010647428.0A CN111898725A (en) | 2020-07-07 | 2020-07-07 | Air conditioning system sensor fault detection method and device and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111898725A true CN111898725A (en) | 2020-11-06 |
Family
ID=73191686
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010647428.0A Pending CN111898725A (en) | 2020-07-07 | 2020-07-07 | Air conditioning system sensor fault detection method and device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111898725A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113362425A (en) * | 2021-06-18 | 2021-09-07 | 中科三清科技有限公司 | Image fusion method and device, electronic equipment and storage medium |
| CN115307670A (en) * | 2022-10-12 | 2022-11-08 | 蘑菇物联技术（深圳）有限公司 | Method, apparatus and medium for locating central air conditioning system anomaly sensors |
| EP4148518A1 (en) | 2021-09-13 | 2023-03-15 | EVER_EST Srl | Kpca-based non-linear chiller fault detection method |
| CN116863419A (en) * | 2023-09-04 | 2023-10-10 | 湖北省长投智慧停车有限公司 | Method and device for lightening target detection model, electronic equipment and medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105825063A (en) * | 2016-03-21 | 2016-08-03 | 北京航空航天大学 | Method for quantitatively selecting test point in design-for-test |
| WO2018072351A1 (en) * | 2016-10-20 | 2018-04-26 | 北京工业大学 | Method for optimizing support vector machine on basis of particle swarm optimization algorithm |
| CN110175682A (en) * | 2019-04-22 | 2019-08-27 | 广东技术师范大学 | A kind of optimization core pivot element analysis fault monitoring method based on Chaos-Particle Swarm Optimization |
-
2020
- 2020-07-07 CN CN202010647428.0A patent/CN111898725A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105825063A (en) * | 2016-03-21 | 2016-08-03 | 北京航空航天大学 | Method for quantitatively selecting test point in design-for-test |
| WO2018072351A1 (en) * | 2016-10-20 | 2018-04-26 | 北京工业大学 | Method for optimizing support vector machine on basis of particle swarm optimization algorithm |
| CN110175682A (en) * | 2019-04-22 | 2019-08-27 | 广东技术师范大学 | A kind of optimization core pivot element analysis fault monitoring method based on Chaos-Particle Swarm Optimization |
Non-Patent Citations (5)
| Title |
|---|
| CHOI, S ETAL.: "Nonlinear Dynamic Process Monitoring Based on Dynamic Kernel PCA", CHEMICAL ENGINEERING SCIENCE, 30 November 2004 (2004-11-30) * |
| 王印松;蔡博;焦阳;朱向伟;: "基于改进类均值核主元分析的控制系统传感器故障检测", 电力科学与工程, no. 09, 28 September 2017 (2017-09-28) * |
| 石怀涛;宋文丽;张珂;周乾;: "基于粒子群优化的核主元分析的故障检测方法", 沈阳建筑大学学报(自然科学版), no. 04, 15 July 2016 (2016-07-15) * |
| 肖应旺等: "一种用于故障监测的优化核主元分析方法", 计算机与应用化学, vol. 36, no. 4, 28 August 2019 (2019-08-28) * |
| 谢宏;李云峰;陈海滨;谭建豪;张爱林;向启均;: "混合粒子群算法优化支持向量机电路故障诊断", 电子测量与仪器学报, no. 03, 15 March 2018 (2018-03-15) * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113362425A (en) * | 2021-06-18 | 2021-09-07 | 中科三清科技有限公司 | Image fusion method and device, electronic equipment and storage medium |
| EP4148518A1 (en) | 2021-09-13 | 2023-03-15 | EVER_EST Srl | Kpca-based non-linear chiller fault detection method |
| CN115307670A (en) * | 2022-10-12 | 2022-11-08 | 蘑菇物联技术（深圳）有限公司 | Method, apparatus and medium for locating central air conditioning system anomaly sensors |
| CN115307670B (en) * | 2022-10-12 | 2023-01-13 | 蘑菇物联技术（深圳）有限公司 | Method, apparatus and medium for locating central air conditioning system anomaly sensors |
| CN116863419A (en) * | 2023-09-04 | 2023-10-10 | 湖北省长投智慧停车有限公司 | Method and device for lightening target detection model, electronic equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111898725A (en) | Air conditioning system sensor fault detection method and device and electronic equipment | |
| Zhang et al. | Building HVAC scheduling using reinforcement learning via neural network based model approximation | |
| Ben-Nakhi et al. | Energy conservation in buildings through efficient A/C control using neural networks | |
| CN109634121B (en) | Multi-parent genetic algorithm air source heat pump multi-objective optimization control method based on radial basis function neural network | |
| US11085663B2 (en) | Building management system with triggered feedback set-point signal for persistent excitation | |
| CN109446028B (en) | Method for monitoring state of refrigeration dehumidifier based on genetic fuzzy C-mean clustering | |
| CN111881505B (en) | Multi-objective optimization transformation decision method for existing building based on GA-RBF algorithm | |
| CN109828578B (en) | Instrument inspection robot optimal route planning method based on YOLOv3 | |
| CN112288164A (en) | Wind power combined prediction method considering spatial correlation and correcting numerical weather forecast | |
| Huang et al. | A new zone temperature predictive modeling for energy saving in buildings | |
| Hou et al. | Nonlinear model predictive control for the space heating system of a university building in Norway | |
| CN111461466A (en) | Heating household valve adjusting method, system and equipment based on L STM time sequence | |
| CN109297498A (en) | A kind of robot path planning method based on improvement multi-target glowworm swarm algorithm | |
| Aqlan et al. | Integrating artificial neural networks and cluster analysis to assess energy efficiency of buildings | |
| CN104898426A (en) | Room temperature loop control method based on gradient descent method and generalized prediction control | |
| CN118043746A (en) | Calibration system and method for calibrating industrial system model using simulation failure | |
| CN105424043A (en) | Motion state estimation method based on maneuver judgment | |
| Sha et al. | A control strategy of heating system based on adaptive model predictive control | |
| CN117075634A (en) | Power distribution network multi-unmanned aerial vehicle scheduling inspection method and device based on improved ant colony algorithm | |
| CN113028610B (en) | Method and device for global optimization and energy-saving control of dynamic load of central air conditioner | |
| Cui et al. | A novel air balancing method for HVAC systems by a full data-driven duct system model | |
| CN116027782A (en) | Distributed similar formation control method considering biased range errors | |
| Drgona et al. | Stripping off the implementation complexity of physics-based model predictive control for buildings via deep learning | |
| CN107590346B (en) | Downscaling correction model based on spatial multi-correlation solution set algorithm | |
| CN117716181A (en) | Method and device for controlling parameter optimization, storage medium and electronic equipment |
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 |