CN115540202A - Air conditioning unit fault prediction method based on multi-dimensional Taylor network - Google Patents
Air conditioning unit fault prediction method based on multi-dimensional Taylor network Download PDFInfo
- Publication number
- CN115540202A CN115540202A CN202211042491.7A CN202211042491A CN115540202A CN 115540202 A CN115540202 A CN 115540202A CN 202211042491 A CN202211042491 A CN 202211042491A CN 115540202 A CN115540202 A CN 115540202A
- Authority
- CN
- China
- Prior art keywords
- fault
- conditioning unit
- air conditioning
- network
- taylor
- 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
Images
Classifications
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
- F24F11/00—Control or safety arrangements
- F24F11/30—Control or safety arrangements for purposes related to the operation of the system, e.g. for safety or monitoring
- F24F11/32—Responding to malfunctions or emergencies
- F24F11/38—Failure diagnosis
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
- F24F11/00—Control or safety arrangements
- F24F11/62—Control or safety arrangements characterised by the type of control or by internal processing, e.g. using fuzzy logic, adaptive control or estimation of values
- F24F11/63—Electronic processing
- F24F11/64—Electronic processing using pre-stored data
Abstract
The invention discloses an air conditioning unit fault prediction method based on a multi-dimensional Taylor network, which is characterized in that on the basis of each fault characteristic corresponding to each type of specified fault in a preset air conditioning unit, a fault static threshold range of an effective characteristic variable in a normal state is obtained through calculation of an exponential weighted moving average control chart, and then a prediction output value based on a multi-dimensional Taylor network model is compared with a static fault threshold, so that the prediction of each type of specified fault in the air conditioning unit is realized. According to the method, the parallel prediction of each fault characteristic is realized by multidimensional fitting of the multidimensional Taylor network BP-MTN model, so that not only is the fitting performance of the original multidimensional Taylor network enhanced, but also the prediction efficiency of the fault characteristic is improved, and further, the method for predicting the fault of the air conditioning unit based on the multidimensional Taylor network can effectively improve the prediction efficiency of the fault of the air conditioning unit.
Description
Technical Field
The invention relates to the technical field of air conditioning unit fault detection, in particular to an air conditioning unit fault prediction method based on a multidimensional Taylor network.
Background
In public buildings, hvac systems are the primary energy consuming equipment, approximately 50-60%. About 42% of the cooling energy consumption and 26% of the maintenance cost in the hvac are caused by equipment failure, and it is estimated that the predictive diagnosis of the hvac failure can reduce the energy consumption by 10% -40%. At present, the energy consumption of China is continuously increased, so that energy-saving and environment-friendly treatment needs to be implemented, and the scientificity and the rationality of the application of the heating, ventilating and air conditioning system are improved. As an important subsystem of a hvac system, air conditioning units are used to condition indoor air to a suitable temperature. And the devices in the air conditioning unit are coupled with each other, and the control of indoor room temperature, humidity and the like is realized through closed-loop control to provide a comfortable environment for users. In this case, if one of the devices in the air conditioning unit fails, a chain reaction is caused, and the transmission and diffusion of the failure are caused. Therefore, in the case of coupled devices, it is necessary to accurately predict the failure of the air conditioning unit.
The prior art has many prediction researches on faults, and the most common faults can be divided into: the method is based on a model fault prediction technology, a data-driven fault prediction technology and a probability statistics fault prediction technology, but the current prediction method is limited in prediction accuracy and low in prediction efficiency in practical application.
Disclosure of Invention
Aiming at the problems, the invention provides an air conditioning unit fault prediction method based on a multi-dimensional Taylor network.
In order to realize the aim of the invention, the invention provides an air conditioning unit fault prediction method based on a multidimensional Taylor network, which comprises the following steps:
s1: based on the fault characteristics corresponding to each fault of the specified type in the preset air conditioning unit, from ASHRAE RP-1312 data set to t 0 Collecting fault characteristic values under l normal operation state time points from the time to the historical time direction to construct an x multidimensional matrix; simultaneously from (t) 0 Collecting fault characteristic values under l normal operation state time points from the moment + t) to the historical time direction to construct a y multi-dimensional matrix so as to form a sample data set (x, y);
s2: carrying out data normalization operation updating on the sample data set (x, y) to obtain a sampleThis data x * Multidimensional matrix and sample data y * A multi-dimensional matrix;
s3: calculating the sample data y * Extracting the sample data y according to the static fault threshold value of each fault feature in the multidimensional matrix * The multidimensional matrix is used as a model estimation state of each fault characteristic in the next t time periodA corresponding label vector;
s4: based on the sample data x * Constructing a multidimensional Taylor network BP-MTN model with same-dimensional input and output according to the dimensionality of a multidimensional matrix, and then sampling data x * The multidimensional matrix is used as a model input characteristic, the multidimensional Taylor network BP-MTN model is input, and the estimated state of the model is outputThen the label vector y * And model estimated stateError calculation is carried out, and finally training of the multi-dimensional Taylor net BP-MTN model is completed according to a back propagation method;
s5: based on the fault characteristics corresponding to each fault of the specified type in the preset air conditioning unit, from ASHRAE RP-1312 data set, from t 0 Collecting l fault characteristic values including normal operation state and fault state time points from the time to the historical time direction to construct an X multidimensional matrix; simultaneously from (t) 0 + t) collecting l fault characteristic values including normal operation state and fault state at time point in historical time direction to construct Y multidimensional matrix, forming data set (X, Y) to be tested, normalizing the data set (X, Y) to be tested to obtain updated data set (X, Y) to be tested * ,Y * );
s6: judging the data set (X) to be tested based on the static fault threshold value of each fault feature calculated in the step s3 and the trained multi-dimensional Taylor net BP-MTN model * ,Y * ) Wherein each fault is characterized by (t) 0 + t) whether the estimated state in the first t time period after the moment is abnormal or not, if so, judging that the corresponding fault feature is an abnormal fault feature, and further judging the fault type corresponding to the abnormal fault feature based on the corresponding relation between the fault feature and the fault type so as to judge whether the fault really occurs or not in the following; otherwise, the air conditioning unit is judged to correspond to the target air conditioning unit (t) 0 + t) the estimated state in the first t period after time is normal.
Further, in the step s3, the sample data y * The calculation formula of the static fault threshold value of each fault characteristic in the multidimensional matrix is as follows:
wherein, UCL represents the upper limit of the threshold value of the static fault characteristics, LCL represents the lower limit of the threshold value of the static fault characteristics, and mu 0 Representing the mean value of the current observation sample, sigma is the standard deviation of the current observation sample, lambda is a smoothing constant between 0 and 1, and L represents a control limit parameter.
Further, the multidimensional Taylor net BP-MTN model comprises: the device comprises an input layer, a data processing layer, an output layer, a full connection layer and an activation function layer, wherein the layers are sequentially connected, and the output of the previous layer is the input of the next layer.
Further, the formula of the weighted summation of the data processing layers is as follows:
wherein, T j Representing the jth output node of the multi-dimensional Taylor network, m representing the degree of the highest expansion term of the data processing layer, N (N, m) representing the term number of a polynomial of N input fault characteristics of the data processing layer after expansion by m powers, q ∈ N (N, m) representing the qth polynomial of the data processing layer after power expansion, w j,q Representing the qth polynomial on the jth output node of the Wittie networkWeight before formula, σ q,i Representing the power on variable x in the qth polynomial.
Further, in step s4, the multidimensional taylor net BP-MTN model performs multidimensional fitting prediction according to the following formula:
wherein, theta j And representing a weight vector after the j node value of the full connection layer enters the activation function, wherein relu is the activation function.
Further, in the step s4, the back propagation algorithm calculates the square of the fitting error according to the following formula:
further, in the step s6,
determining the data set (X) to be tested * ,Y * ) Wherein each fault is characterized by (t) 0 + t) the process of whether there is an abnormality in the estimated state in the first t time period after the time comprises: and performing statistical identification by marking fault nodes and fault types.
Firstly, setting each fault feature as a fault feature node, setting a node value to be 0 when the fault feature is in a normal state, setting the node value to be 1 when the fault feature is abnormal, then forming a fault feature statistical vector [1,0,1, … …,0] through 0,1, wherein the vector arrangement sequence is consistent with the node sorting sequence, and simultaneously coding the fault type through 1,0, namely judging the fault type through comparing the state value statistics of the fault node with the fault coding.
Further, each fault of the specified type in the preset air conditioning unit includes: the air return clip has the faults of fixed speed, complete damage of an air return fan, full-closed position failure of an outdoor air valve clip, full-open position failure of a cooling coil valve clip and fixed angle failure of the cooling coil valve clip;
the return air card includes at the fault characteristic that fixed speed trouble corresponds: the energy of a return air fan, the air supply temperature of an air conditioner, the air supply rate and the return air rate;
the fault characteristics corresponding to the complete damage fault of the return air fan comprise: the air supply temperature, the air return rate and the air mixing temperature of the air conditioner;
the fault characteristics that the outdoor air blast gate card dies at the full closed position fault correspondence include: outdoor air flow rate, return air fan energy and supply air temperature;
the cooling coil valve card is dead in the fault characteristics that full open position trouble corresponds include: outdoor air flow rate, air supply temperature and air return temperature;
the cooling coil valve card dies the fault characteristics that correspond at fixed angle trouble and include: outdoor air flow rate and return air rate.
Compared with the prior art, the invention has the following beneficial technical effects:
the invention designs a fault prediction method of an air conditioning unit based on a multidimensional Taylor network, which is based on each fault characteristic corresponding to each type of specified faults in a preset air conditioning unit respectively, calculates and obtains a fault threshold value range of an effective characteristic variable in a normal state, and realizes the prediction of each type of specified faults in the air conditioning unit by comparing a multi-dimensional Taylor network prediction output value with the fault threshold value range; therefore, the method and the device can effectively improve the efficiency of predicting the faults of the air conditioning unit.
Drawings
FIG. 1 is a schematic flow chart of a multi-dimensional Taylor network-based air conditioning unit fault prediction method according to an embodiment;
FIG. 2 is a network architecture diagram of one embodiment of a multidimensional Taylor network adding a fully connected layer and an activation function layer;
FIG. 3 is a fault signature node identification statistical graph of one embodiment;
fig. 4 (a), (b), (c) are waveform diagrams illustrating the principle diagnostic waveforms of the cooling coil valve stuck in the fully open position according to one embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is explicitly and implicitly understood by one skilled in the art that the embodiments described herein can be combined with other embodiments.
The invention designs a fault prediction method of an air conditioning unit based on a multidimensional Taylor network, which is used for realizing the prediction of various types of specified faults in the air conditioning unit based on the fault characteristics corresponding to the specified faults in the preset air conditioning unit; in practical applications, as shown in fig. 1, the following steps are performed in real time.
And 2, carrying out data normalization operation on the multidimensional matrixes x and y according to the following formula:
wherein: min (x) i ) Represents the minimum value, max (x) in the ith fault characteristic state vector input in the multi-dimensional matrix of the sample data set x i ) Representing the maximum value of the ith fault characteristic state vector input in the sample data set x multidimensional matrix; min (y) i ) Represents the minimum value of the label value in the ith fault characteristic state vector in the y multidimensional matrix of the sample data set, max (y) i ) Representing the maximum value of the label value in the ith dimension fault characteristic state vector in the y multidimensional matrix of the sample data set; thus obtaining normalized sample data x * 、y * As a network input; then step 3 is entered.
Step 3, calculating y * The fault threshold value of each fault characteristic in the multidimensional matrix is obtained according to the following formula:
wherein, UCL is the upper limit of the static fault characteristic threshold, LCL is the lower limit of the static fault characteristic threshold, mu 0 The mean value of the current observation sample, sigma is the standard deviation of the current observation sample, lambda is a smoothing constant between 0 and 1, and L is a control limit parameter, which is usually set to 1.5; then step 4 is entered.
And 4, building a BP-MTN model according to the dimension of the sample input data x, wherein the MTN structure is divided into an input layer, a data processing layer and an output layer, wherein the node of the input and output layer is equal to the dimension of x, and the data processing layer is used for realizing the weighted summation of each power product term of the input variable according to the following formula:
wherein m represents the times of the highest expansion term of the data processing layer of the multidimensional Taylor network, N (N, m) represents the term number of a polynomial after N input fault characteristics of the data processing layer in the multidimensional Taylor network are expanded by m powers, q is an integer N (N, m) represents a q-th polynomial after the data processing layer is expanded by m powers, and T is j Representing the jth output node, w, of the multidimensional Taylor network j,q Representing the weight, σ, before the qth polynomial on the jth output node of the Wittie network q,i Representing the variable x in the qth polynomial * The power of (a) or (b); then step 5 is entered.
And 5, as shown in fig. 2, adding a full connection layer and an activation function layer after the multidimensional taylor net structure built in the step 4, wherein multidimensional fitting prediction is according to the following formula:
wherein, theta j Representing a weight vector after the j node value of the full connection layer enters the activation function, wherein relu is the activation function; then step 6 is entered.
And 6, carrying out model training according to a back propagation algorithm principle, and calculating the square of a fitting error according to the following formula:
taking the output of the j node as an example to perform back propagation, the network related weight gradient is calculated by the following formula:
calculating the average gradient of each weight value from the gradient vector obtained by the formula (6)Updating the weight value according to the following formula
Then step 7 is entered.
Step 7, judging whether each fault characteristic respectively corresponds to the abnormal state in the prediction state of the next time of the current time according to the normal threshold interval of each fault characteristic data, if so, judging each fault characteristic with the abnormal prediction state as each abnormal fault characteristic, and further judging that each specified type fault corresponding to each abnormal fault characteristic exists in the prediction state of the next time of the current time corresponding to the air conditioning unit by combining the corresponding relation between each specified type fault in the preset air conditioning unit and each related fault characteristic; otherwise, judging that the prediction state of the air conditioning unit corresponding to the next time of the current time is normal. As shown in fig. 3, n identification statistical nodes corresponding to the input fault features are established, so as to distinguish the corresponding relationship between the fault features and the fault types, and perform the fault type prediction.
In practical application, the specified various types of faults include a fault that a return air card is clamped at a fixed speed, a fault that a return air fan is completely damaged, a fault that an outdoor air valve is clamped at a full-closed position, a fault that a cooling coil valve is clamped at a full-open position, and a fault that the cooling coil valve is clamped at a fixed angle.
Wherein, each fault characteristic that the return air card corresponds at the fixed speed trouble includes: return air fan energy, air conditioner air supply temperature, air supply rate, return air rate.
Each fault characteristic that the return air fan damages the trouble completely and corresponds includes: air supply temperature, air return rate and air mixing temperature of the air conditioner.
The outdoor air blast gate card dies each fault characteristic that corresponds at the position trouble of totally closing includes: outdoor air flow rate, return air fan energy, and supply air temperature.
The cooling coil valve card is died in each trouble characteristic that full open position trouble corresponds includes: outdoor air flow rate, air supply temperature and air return temperature.
The cooling coil valve card is dead in each trouble characteristic that fixed angle trouble corresponds includes: outdoor air flow rate, return air rate.
In practical application, based on each fault characteristic corresponding to each type of fault in the air conditioning unit, the prediction of each type of fault in the air conditioning unit is realized; according to the air conditioning unit fault prediction method based on the multi-dimensional Taylor network, the fault threshold value range of the effective characteristic variable in the normal state is calculated and obtained based on the fault characteristics corresponding to the faults of each type specified in the preset air conditioning unit, and the prediction of the faults of each type specified in the air conditioning unit is realized by comparing the prediction output value of the multi-dimensional Taylor network with the fault threshold value range, for example, the fault diagnosis principle waveform diagrams of the cooling coil pipe valve stuck at the full-open position are shown in the diagrams of fig. 4 (a), (b) and (c). Compared with other data-driven fault prediction technologies, the multidimensional Taylor network structure can express complex nonlinear functions by mathematical combination of series terms, can fit a nonlinear system under conventional conditions without carrying out a large amount of network training, and has the advantages of simple network structure, strong network generalization capability, certain interpretability and stronger approximation performance; therefore, the air conditioning unit fault prediction method can effectively improve the prediction efficiency of the air conditioning unit fault.
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
It should be noted that the terms "first \ second \ third" referred to in the embodiments of the present application merely distinguish similar objects, and do not represent a specific ordering for the objects, and it should be understood that "first \ second \ third" may exchange a specific order or sequence when allowed. It should be understood that "first \ second \ third" distinct objects may be interchanged under appropriate circumstances such that the embodiments of the application described herein may be implemented in an order other than those illustrated or described herein.
The terms "comprising" and "having" and any variations thereof in the embodiments of the present application are intended to cover non-exclusive inclusions. For example, a process, method, apparatus, product, or device that comprises a list of steps or modules is not limited to the listed steps or modules but may alternatively include other steps or modules not listed or inherent to such process, method, product, or device.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent application shall be subject to the appended claims.
Claims (8)
1. A multi-dimensional Taylor network-based air conditioning unit fault prediction method is characterized by comprising the following steps:
s1: based on the fault characteristics corresponding to each fault of the specified type in the preset air conditioning unit, from ASHRAE RP-1312 data set to t 0 Collecting fault characteristic values under l normal operation state time points from the time to the historical time direction to construct an x multidimensional matrix; simultaneously from (t) 0 Collecting fault characteristic values under l normal operation state time points from the moment + t) to the historical time direction to construct a y multi-dimensional matrix so as to form a sample data set (x, y);
s2: carrying out data normalization operation updating on the sample data set (x, y) to obtain sample data x * Multidimensional matrix and sample data y * A multi-dimensional matrix;
s3: calculating the sample data y * Extracting the sample data y according to the static fault threshold value of each fault feature in the multidimensional matrix * The multidimensional matrix is used as a model estimation state of each fault characteristic in the next t time periodA corresponding tag vector;
s4: based on the sample data x * Establishing a multidimensional Taylor network BP-MTN model with same-dimensional input and output according to the dimensionality of the multidimensional matrix, and then sampling data x * The multidimensional matrix is used as a model input characteristic, the multidimensional Taylor network BP-MTN model is input, and the estimated state of the model is outputThen for the label vector y * And model estimated stateError calculation is carried out, and finally training of the multi-dimensional Taylor net BP-MTN model is completed according to a back propagation method;
s5: based on the fault characteristics corresponding to each fault of the specified type in the preset air conditioning unit, from ASHRAE RP-1312 data set, from t 0 Collecting l fault characteristic values including normal operation state and fault state time points from the time to the historical time direction to construct an X multidimensional matrix; at the same time from (t) 0 + t) collecting l fault characteristic values including normal operation state and fault state at time point in historical time direction to construct Y multidimensional matrix, forming data set (X, Y) to be tested, normalizing the data set (X, Y) to be tested to obtain updated data set (X, Y) to be tested * ,Y * );
s6: judging the data set (X) to be tested based on the static fault threshold value of each fault characteristic calculated in the step s3 and the trained multi-dimensional Taylor net BP-MTN model * ,Y * ) Wherein each fault is characterized by (t) 0 + t) whether the estimated state in the first t time period after the moment is abnormal or not, if so, judging the corresponding fault feature as the abnormal fault feature, and further judging the fault type corresponding to the abnormal fault feature based on the corresponding relation between the fault feature and the fault type so as to judge whether the fault is ensured or not in the following processActually taking place; otherwise, the air conditioning unit is judged to correspond to the target air conditioning unit (t) 0 + t) the estimated state in the first t period after time is normal.
2. The multi-dimensional Taylor network-based air conditioning unit fault prediction method according to claim 1, wherein in the step s3, the sample data y * The calculation formula of the static fault threshold value of each fault characteristic in the multidimensional matrix is as follows:
wherein, UCL represents the upper limit of the threshold value of the static fault characteristics, LCL represents the lower limit of the threshold value of the static fault characteristics, and mu 0 Representing the mean value of the current observation sample, sigma is the standard deviation of the current observation sample, lambda is a smoothing constant between 0 and 1, and L represents a control limit parameter.
3. The multi-dimensional Taylor network-based air conditioning unit fault prediction method as claimed in claim 2,
the multidimensional Taylor net BP-MTN model comprises: the device comprises an input layer, a data processing layer, an output layer, a full connection layer and an activation function layer, wherein the layers are sequentially connected, and the output of the previous layer is the input of the next layer.
4. The multi-dimensional Taylor network-based air conditioning unit fault prediction method as claimed in claim 3,
the formula of the weighted summation of the data processing layers is as follows:
wherein, T j Representing the jth output node of the multi-dimensional Taylor network, m representing the number of times of the highest expansion item of the data processing layer, and N (N, m) representing N input fault characteristics of the data processing layerThe term number of the polynomial after the m-power expansion is carried out, q belongs to N (N, m) and represents the q-th polynomial after the power expansion of the data processing layer, w j,q Representing the weight, σ, before the qth polynomial on the jth output node of the Wittie network q,i Representing the power on variable x in the qth polynomial.
5. The multi-dimensional Taylor network-based air conditioning unit fault prediction method as claimed in claim 4,
in step s4, the multidimensional Taylor net BP-MTN model carries out multidimensional fitting prediction according to the following formula:
wherein, theta j And representing a weight vector after the j node value of the full connection layer enters the activation function, wherein relu is the activation function.
7. the multi-dimensional Taylor network-based air conditioning unit fault prediction method according to claim 6, wherein in the step s6,
determining the data set (X) to be tested * ,Y * ) Wherein each fault is characterized by (t) 0 The process of judging whether the estimated state in the first t time period after the moment + t) is abnormal or not comprises the following steps: and performing statistical identification by marking fault nodes and fault types.
8. The multi-dimensional Taylor network-based air conditioning unit fault prediction method as claimed in claim 7,
each fault of the specified type in the preset air conditioning unit comprises: the return air card has the faults of fixed speed, the return air fan is completely damaged, the outdoor air valve is stuck at a full-closed position, the cooling coil valve is stuck at a full-open position, and the cooling coil valve is stuck at a fixed angle;
the return air card includes at the fault characteristic that fixed speed trouble corresponds: the energy of a return air fan, the air supply temperature of an air conditioner, the air supply speed and the return air speed;
the fault characteristics corresponding to the complete damage fault of the return air fan comprise: the air supply temperature, the air return rate and the air mixing temperature of the air conditioner;
the fault characteristics that the outdoor air blast gate card dies at the full closed position fault correspondence include: outdoor air flow rate, return air fan energy and supply air temperature;
the cooling coil valve card is dead in the fault characteristics that full open position trouble corresponds include: outdoor air flow rate, air supply temperature and air return temperature;
the cooling coil valve card is dead in the fault characteristics that fixed angle fault corresponds including: outdoor air flow rate and return air rate.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211042491.7A CN115540202A (en) | 2022-08-29 | 2022-08-29 | Air conditioning unit fault prediction method based on multi-dimensional Taylor network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211042491.7A CN115540202A (en) | 2022-08-29 | 2022-08-29 | Air conditioning unit fault prediction method based on multi-dimensional Taylor network |
Publications (1)
Publication Number | Publication Date |
---|---|
CN115540202A true CN115540202A (en) | 2022-12-30 |
Family
ID=84726602
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202211042491.7A Pending CN115540202A (en) | 2022-08-29 | 2022-08-29 | Air conditioning unit fault prediction method based on multi-dimensional Taylor network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115540202A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117648612A (en) * | 2024-01-30 | 2024-03-05 | 上海移视网络科技有限公司 | Parallel battery pack fault detection method, device, electronic equipment and storage medium |
-
2022
- 2022-08-29 CN CN202211042491.7A patent/CN115540202A/en active Pending
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117648612A (en) * | 2024-01-30 | 2024-03-05 | 上海移视网络科技有限公司 | Parallel battery pack fault detection method, device, electronic equipment and storage medium |
CN117648612B (en) * | 2024-01-30 | 2024-04-12 | 上海移视网络科技有限公司 | Parallel battery pack fault detection method, device, electronic equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10747187B2 (en) | Building management system with voting-based fault detection and diagnostics | |
Wang et al. | Fault detection based on Bayesian network and missing data imputation for building energy systems | |
US11579952B2 (en) | Machine-learning based optimization of data center designs and risks | |
Zhou et al. | A comparison study of basic data-driven fault diagnosis methods for variable refrigerant flow system | |
CN111881505B (en) | Multi-objective optimization transformation decision method for existing building based on GA-RBF algorithm | |
CN114484731A (en) | Method and device for diagnosing faults of central air conditioner based on stacking fusion algorithm | |
CN111140986A (en) | Operating state detection method and device of air conditioning system, storage medium and air conditioner | |
US11102083B2 (en) | Machine-learning based optimization of data centers using supplemental indicators | |
CN113240184B (en) | Building space unit cold load prediction method and system based on federal learning | |
CN110543932A (en) | air conditioner performance prediction method and device based on neural network | |
CN112283876A (en) | Air conditioner fault prediction method and air conditioner | |
CN115540202A (en) | Air conditioning unit fault prediction method based on multi-dimensional Taylor network | |
Guo et al. | Intelligent model based fault detection and diagnosis for HVAC system using statistical machine learning methods | |
CN111156657A (en) | Air conditioner frosting state determining method and device | |
CN113405223A (en) | Cold machine number control method based on GRNN and control system thereof | |
Li et al. | Support vector machine based fault detection and diagnosis for HVAC systems | |
Feng et al. | A fully distributed voting strategy for AHU fault detection and diagnosis based on a decentralized structure | |
AU2021100960A4 (en) | Artificial Intelligence Based Cooling System for Managing the Energy Efficiency | |
Li et al. | Simulated annealing wrapped generic ensemble fault diagnostic strategy for VRF system | |
Wei et al. | Fault detection and diagnosis for variable-air-volume systems using combined residual, qualitative and quantitative techniques | |
Adnan et al. | Modeling baseline electrical energy use of chiller system by artificial neural network | |
Beghi et al. | Load forecasting for the efficient energy management of HVAC systems | |
CN113268913A (en) | Intelligent building air conditioner cooling machine system operation optimization method based on PSO-ELM algorithm | |
Aoki et al. | Smart grid real-time pricing optimization control with simulated annealing algorithm for office building air-conditioning facilities | |
Wang et al. | Fault detection and diagnosis of HVAC system based on federated learning |
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 |