CN114858467B - Diesel engine anti-noise and cross-noise domain fire diagnosis method and system - Google Patents
Diesel engine anti-noise and cross-noise domain fire diagnosis method and system Download PDFInfo
- Publication number
- CN114858467B CN114858467B CN202210589677.8A CN202210589677A CN114858467B CN 114858467 B CN114858467 B CN 114858467B CN 202210589677 A CN202210589677 A CN 202210589677A CN 114858467 B CN114858467 B CN 114858467B
- Authority
- CN
- China
- Prior art keywords
- noise
- module
- diesel engine
- signal
- convolution
- 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
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M15/00—Testing of engines
- G01M15/04—Testing internal-combustion engines
- G01M15/11—Testing internal-combustion engines by detecting misfire
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M15/00—Testing of engines
- G01M15/04—Testing internal-combustion engines
- G01M15/12—Testing internal-combustion engines by monitoring vibrations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
- G06F2218/04—Denoising
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Chemical & Material Sciences (AREA)
- Combustion & Propulsion (AREA)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Combined Controls Of Internal Combustion Engines (AREA)
Abstract
The invention provides a diesel engine anti-noise and cross-noise domain fire diagnosis method and system, comprising the following steps: collecting vibration signals of a cylinder cover of a diesel engine on line; based on a noise superposition principle, a residual convolution preprocessing module is designed to perform noise reduction preprocessing on an original vibration signal measured by a diesel engine cylinder, and a residual loss construction loss function is utilized for model training; designing a multi-scale convolution module to extract fault characteristics of different time scales from the preprocessed signals; extracting signal dependent characteristics from the signal processed in the step S3 by using LSTM; and constructing a diesel engine fire diagnosis model by using a Keras packet under a TensorFlow framework, training, and evaluating the performance of the module according to the diagnosis result and the actual comparison. According to the invention, residual convolution preprocessing and a multi-scale convolution long-term and short-term memory network are integrated, so that the essential characteristics of the fire of the diesel engine are fully excavated, and intelligent diagnosis of the fire of the diesel engine under strong noise and different noise domains is realized.
Description
Technical Field
The invention relates to the field of diesel engine fault diagnosis, in particular to a diesel engine anti-noise and cross-noise-domain fire diagnosis method and system, and more particularly relates to a diesel engine anti-noise and cross-noise-domain fire diagnosis method and system integrating residual convolution preprocessing and a multi-scale convolution long-short-term memory network.
Background
The multi-cylinder diesel engine works stably and can obtain enough power. Compared with the gasoline engine, the diesel engine has the obvious advantages of long service life, economy, durability, low speed, large torque, safety, environmental protection and the like. Therefore, it has been widely used in the fields of engineering machinery, automobile industry, ship machinery, electric power industry, agricultural machinery, etc. Misfires are common faults of diesel engines, mainly caused by faults of electrical control systems and faults of mechanical components. Electronic control system faults, including loss or inaccuracy of sensor signals, control unit control signal faults or no signal output, ignition faults caused by damage to spark plugs or ignition coils, injector injection faults, and circuit connection faults. The mechanical failure is mainly caused by insufficient cylinder pressure, such as loose valve closing, leakage and the like. The fire may cause severe vibration, insufficient power, weak acceleration, and high fuel consumption. Therefore, it is of great importance to monitor the running state of the engine on line and take corresponding measures.
Patent document CN103032190B (application number 2012105725412) discloses a method and apparatus for detecting a misfire of a diesel engine based on a rail pressure signal, comprising: collecting instantaneous rail pressures respectively corresponding to a starting tooth and an ending tooth of a current cylinder; calculating a rail pressure drop value generated by the instantaneous rail pressure of the start tooth and the end tooth of each cylinder; calculating rail pressure drop standard values from rail pressure drop values of all cylinders; dividing the rail pressure drop value of each cylinder with the rail pressure drop standard value to obtain a rail pressure drop proportional value of each cylinder; judging that the rail pressure drop value proportion value of each cylinder is smaller than a preset relative fire threshold value, and judging that the cylinder fires when the rail pressure drop value of the corresponding cylinder is smaller than a preset absolute fire threshold value; and otherwise, judging that the fire is not caused.
Patent document CN102980777B (application number 2012105627710) discloses a method and apparatus for detecting a misfire of a diesel engine based on single cylinder angular acceleration, comprising: collecting the instantaneous rotation speeds of crankshaft teeth corresponding to the starting teeth and the ending teeth of the current cylinder; calculating single-cylinder angular acceleration of each cylinder from the instantaneous rotational speed of the crankshaft teeth; calculating a single-cylinder angular acceleration standard value from the single-cylinder angular accelerations of all cylinders; dividing the single-cylinder angular acceleration of each cylinder with the angular acceleration standard value to obtain a single-cylinder angular acceleration proportional value of each cylinder; judging that the cylinder is in fire when the single cylinder angular acceleration ratio value of each cylinder is smaller than a preset relative fire threshold value and the single cylinder angular acceleration of the corresponding cylinder is smaller than a preset absolute fire threshold value; otherwise, no fire occurs.
The above patent needs to set the fire threshold value when diagnosing the fire, is difficult to avoid the diagnosis error caused by artificial experience, and because the measurement signal is generally interfered by various noises, the fire diagnosis precision of the diesel engine under strong noise and the generalization performance among different noise levels are still limited in practice.
Disclosure of Invention
Aiming at the defects in the prior art, the invention aims to provide a diesel engine anti-noise and cross-noise-domain fire diagnosis method and system.
The invention provides a diesel engine anti-noise and cross-noise domain fire diagnosis method, which comprises the following steps:
step S1: collecting vibration signals of a cylinder cover of a diesel engine on line;
step S2: based on a noise superposition principle, a residual convolution preprocessing module is designed to perform noise reduction preprocessing on an original vibration signal measured by a diesel engine cylinder, and a residual loss construction loss function is utilized for model training;
step S3: designing a multi-scale convolution module to extract fault characteristics of different time scales from the preprocessed signals;
step S4: extracting signal dependent characteristics from the signal processed in the step S3 by using LSTM;
step S5: and constructing a diesel engine fire diagnosis model by using a Keras packet under a TensorFlow framework, training, and evaluating the performance of the module according to the diagnosis result and the actual comparison.
Preferably, the signal after noise reduction in the step S2 is the original signal minus the residual convolution preprocessing module output signal.
Preferably, the loss function in the step S2 is:
wherein loss is β For the constructed loss function, n is the number of samples transferred to the model per iteration, β is the preset noise template, output i Outputting a noise signal obtained by the ith sample signal after the residual convolution preprocessing module, and outputting beta i Representing the random noise signal carried by the i-th sample signal.
Preferably, the multi-scale convolution module comprises a plurality of convolution layer modules, each convolution layer module comprising a different branch and a different size convolution kernel.
Preferably, a plurality of the convolution layer modules are connected in series, and fault features of different time scales are extracted.
According to the invention, a diesel engine anti-noise and cross-noise-domain fire diagnosis system comprises:
module M1: collecting vibration signals of a cylinder cover of a diesel engine on line;
module M2: based on a noise superposition principle, carrying out noise reduction pretreatment on an original vibration signal measured by a diesel engine cylinder through a residual convolution pretreatment module, and constructing a loss function by utilizing residual loss for model training;
module M3: extracting fault characteristics of different time scales from the preprocessed signals by utilizing a multi-scale convolution module;
module M4: extracting signal dependent characteristics from the signal processed by the module M3 by using LSTM;
module M5: and constructing a diesel engine fire diagnosis model by using a Keras packet under a TensorFlow framework, training, and evaluating the performance of the module according to the diagnosis result and the actual comparison.
Preferably, the signal after noise reduction by the module M2 is the original signal minus the residual convolution preprocessing module output signal.
Preferably, the loss function in the module M2 is:
wherein loss is β For the constructed loss function, n is the number of samples transferred to the model per iteration, β is the preset noise template, output i Outputting a noise signal obtained by the ith sample signal after the residual convolution preprocessing module, and outputting beta i Representing the random noise signal carried by the i-th sample signal.
Preferably, the multi-scale convolution module comprises a plurality of convolution layer modules, each convolution layer module comprising a different branch and a different size convolution kernel.
Preferably, a plurality of the convolution layer modules are connected in series, and fault features of different time scales are extracted.
Compared with the prior art, the invention has the following beneficial effects:
1. the invention designs a residual convolution preprocessing module to process an original vibration signal measured by a diesel engine cylinder, and the residual loss constructs a new loss function for model training, thereby providing a basis for the subsequent extraction of the essential fault of the fire failure of the diesel engine;
2. the invention designs a multi-scale convolution module to realize multi-scale feature extraction so as to enhance the robustness of the model in time scale when learning the fire fault feature, and simultaneously utilizes LSTM to extract correlation features so as to further improve the adaptability performance of anti-noise and noise domain;
3. according to the invention, the residual convolution preprocessing and the multi-scale convolution long-term and short-term memory network are integrated, the essential characteristics of the diesel engine fire are fully excavated, the intelligent diagnosis of the diesel engine fire under strong noise and different noise domains is realized, the corresponding operation and maintenance decision is adopted later, and the automation and intelligent level of the diesel engine is improved.
Drawings
Other features, objects and advantages of the present invention will become more apparent upon reading of the detailed description of non-limiting embodiments, given with reference to the accompanying drawings in which:
FIG. 1 is a flow chart of an implementation of the diesel anti-noise and cross-noise domain misfire diagnostic method presented by the present invention;
FIG. 2 is a network architecture diagram of the diesel engine anti-noise and cross-noise domain misfire diagnostic model presented by the present invention;
Detailed Description
The present invention will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the present invention, but are not intended to limit the invention in any way. It should be noted that variations and modifications could be made by those skilled in the art without departing from the inventive concept. These are all within the scope of the present invention.
Aiming at the problems that the fire threshold value is required to be set in the current diesel engine fire diagnosis, the fire diagnosis precision under the condition of strong noise and the generalization performance among different noise levels are limited, the invention provides a diesel engine anti-noise and cross-noise domain fire diagnosis method and system which are used for fusing residual convolution preprocessing and a multi-scale convolution long-short-term memory network. According to the invention, the residual convolution preprocessing module is designed to process the original vibration signal measured by the diesel engine cylinder, the multi-scale convolution module is designed to realize multi-scale feature extraction, the robustness of the model in time scale when the fire fault feature is learned is enhanced, and the correlation feature is extracted by utilizing the LSTM so as to further improve the adaptability performance of anti-noise and noise domain. The invention integrates residual convolution preprocessing and a multi-scale convolution long-term and short-term memory network, fully excavates the essential characteristics of the diesel engine fire, and can be used for intelligent diagnosis of the diesel engine fire under strong noise and different noise domains through correlation analysis.
Example 1:
the invention discloses a diesel engine anti-noise and cross-noise domain fire diagnosis method, which comprises the following steps:
step S1: collecting vibration signals of a cylinder cover of a diesel engine on line, and constructing a data segment input model for processing every 1024 points;
step S2: based on a noise superposition principle, a seven-layer residual convolution preprocessing module is designed to perform noise reduction preprocessing on an original vibration signal measured by a diesel engine cylinder, the output of a convolution block is subtracted from the original signal to obtain a denoised signal, a residual loss construction loss function is used for model training, and the loss function is as follows:
wherein loss is β For the constructed loss function, n is the number of samples transferred to the model per iteration, β is the preset noise template, output i Outputting a noise signal obtained by the ith sample signal after the residual convolution preprocessing module, and outputting beta i Representing the random noise signal carried by the i-th sample signal.
Step S3: and designing a multi-scale convolution module to extract fault characteristics of different time scales from the preprocessed signals, wherein the multi-scale convolution module comprises a plurality of convolution layer modules, and each convolution layer module comprises different branches and convolution kernels of different sizes. In the embodiment, the number of the convolution layer modules is preferably four, and the four convolution layer modules are mutually connected in series and are used for extracting different time scale features in the signal so as to enhance the robustness of the time scale when the model learns the fire fault features;
step S4: extracting signal dependency characteristics from the signal processed in the step S3 by utilizing LSTM, and further improving anti-noise and cross-noise domain adaptability;
step S5: and constructing a diesel engine fire diagnosis model by using a Keras packet under a TensorFlow framework, training, and evaluating the performance of the module according to the diagnosis result and the actual comparison.
The diesel engine fire diagnosis model integrates residual convolution preprocessing and a multi-scale convolution long-short-term memory network, fully excavates the essential characteristics of the diesel engine fire, realizes intelligent diagnosis of the diesel engine fire under strong noise and different noise domains, is beneficial to follow-up corresponding operation and maintenance decision, and improves the automation and intelligent level of the diesel engine.
The invention also discloses a diesel engine anti-noise and cross-noise domain fire diagnosis system, which comprises:
module M1: collecting vibration signals of a cylinder cover of a diesel engine on line, and constructing a data segment input model for processing every 1024 points;
module M2: based on the noise superposition principle, the original vibration signal measured by the diesel engine cylinder is subjected to noise reduction pretreatment by a residual convolution pretreatment module, and a residual loss construction loss function is used for model training, wherein the loss function is as follows:
wherein loss is β For the constructed loss function, n is the number of samples transferred to the model per iteration, β is the preset noise template, output i Outputting a noise signal obtained by the ith sample signal after the residual convolution preprocessing module, and outputting beta i Representing the random noise signal carried by the i-th sample signal.
Module M3: extracting fault characteristics of different time scales from the preprocessed signals by utilizing a multi-scale convolution module; the multi-scale convolution module comprises a plurality of convolution layer modules, each convolution layer module comprises convolution kernels with different branches and different sizes, and the convolution layer modules are mutually connected in series to extract fault features with different time scales
Module M4: extracting signal dependent characteristics from the signal processed by the module M3 by using LSTM;
module M5: and constructing a diesel engine fire diagnosis model by using a Keras packet under a TensorFlow framework, training, and evaluating the performance of the module according to the diagnosis result and the actual comparison.
Example 2:
example 2 is a modification of example 1.
Referring to fig. 1 to 2, the present invention provides a diesel engine anti-noise and cross-noise domain misfire diagnosis method, comprising the steps of:
step S1: collecting vibration signals of a cylinder cover of a diesel engine on line, and constructing a data segment input model for processing every 1024 points;
step S2: based on the noise superposition principle, a residual convolution preprocessing module is designed to perform noise reduction preprocessing on an original vibration signal measured by a diesel engine cylinder, and a new loss function is constructed by utilizing residual loss for model training:
wherein loss is β For the constructed loss function, n is the batch size, i.e., the number of samples transferred to the model per iteration, β is the preset noise template, output i And outputting a noise signal obtained by the ith sample signal after the ith sample signal passes through a residual convolution preprocessing module, wherein beta represents a random noise signal carried by the ith sample signal.
Step S3: four multi-scale convolution modules are designed to realize multi-scale feature extraction so as to enhance the robustness of the model in time scale when learning the fire fault feature;
step S4: extracting correlation features by using a single-layer LSTM to further improve anti-noise and noise domain adaptability;
step S5: and constructing a diesel engine fire diagnosis model by using a Keras packet under a TensorFlow framework, training, and evaluating the performance of the model according to the diagnosis result and the actual comparison.
The model is tested and evaluated, and the average accuracy of the model is up to 97% under the condition of strong noise (-10 dB signal to noise ratio) of four data sets under different working conditions. Meanwhile, training is carried out by using a data set with the signal-to-noise ratio of-10 dB under the same working condition, and cross-noise domain diagnosis is respectively carried out by using test sets with the signal-to-noise ratios of-8 dB, -6dB, -4dB, -2dB,0dB,2dB,4dB,6dB,8dB and 10dB, and the accuracy of the model is 97.851%, 97.851%, 97.460%, 97.851%, 97.070%, 97.460%, 97.460% and 97.460% respectively. The result shows that the intelligent diagnosis model for the diesel engine fire is fused with residual convolution preprocessing and a multi-scale convolution long-term and short-term memory network, so that the essential characteristics of the diesel engine fire are fully excavated, the intelligent diagnosis model can be used for intelligent diagnosis of the diesel engine fire under strong noise and different noise domains, corresponding operation and maintenance decisions can be adopted later, and the automation and intelligent level of the diesel engine can be improved.
Those skilled in the art will appreciate that the invention provides a system and its individual devices, modules, units, etc. that can be implemented entirely by logic programming of method steps, in addition to being implemented as pure computer readable program code, in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers, etc. Therefore, the system and various devices, modules and units thereof provided by the invention can be regarded as a hardware component, and the devices, modules and units for realizing various functions included in the system can also be regarded as structures in the hardware component; means, modules, and units for implementing the various functions may also be considered as either software modules for implementing the methods or structures within hardware components.
The foregoing describes specific embodiments of the present invention. It is to be understood that the invention is not limited to the particular embodiments described above, and that various changes or modifications may be made by those skilled in the art within the scope of the appended claims without affecting the spirit of the invention. The embodiments of the present application and features in the embodiments may be combined with each other arbitrarily without conflict.
Claims (6)
1. A diesel engine anti-noise and cross-noise domain misfire diagnostic method, comprising:
step S1: collecting vibration signals of a cylinder cover of a diesel engine on line;
step S2: based on a noise superposition principle, a residual convolution preprocessing module is designed to perform noise reduction preprocessing on an original vibration signal measured by a diesel engine cylinder, and a residual loss construction loss function is utilized for model training;
step S3: designing a multi-scale convolution module to extract fault characteristics of different time scales from the preprocessed signals;
step S4: extracting signal dependent characteristics from the signal processed in the step S3 by using LSTM;
step S5: constructing a diesel engine fire diagnosis model by using a Keras packet under a TensorFlow framework, training, and evaluating the performance of a module according to the diagnosis result and the actual comparison;
the signal after noise reduction in the step S2 is the original signal minus the signal output by the residual convolution preprocessing module;
the loss function in the step S2 is:
wherein loss is β For the constructed loss function, n is the number of samples transferred to the model per iteration, β is the preset noise template, output i Outputting a noise signal obtained by the ith sample signal after the residual convolution preprocessing module, and outputting beta i Representing the random noise signal carried by the i-th sample signal.
2. The diesel engine anti-noise and cross-noise domain misfire diagnostic method as recited in claim 1 wherein: the multi-scale convolution module includes a plurality of convolution layer modules, each convolution layer module including different branches and different sizes of convolution kernels.
3. The diesel anti-noise and cross-noise domain misfire diagnostic method as recited in claim 2 wherein: and the convolution layer modules are mutually connected in series to extract fault characteristics of different time scales.
4. A diesel anti-noise and cross-noise domain misfire diagnostic system, comprising:
module M1: collecting vibration signals of a cylinder cover of a diesel engine on line;
module M2: based on a noise superposition principle, carrying out noise reduction pretreatment on an original vibration signal measured by a diesel engine cylinder through a residual convolution pretreatment module, and constructing a loss function by utilizing residual loss for model training;
module M3: extracting fault characteristics of different time scales from the preprocessed signals by utilizing a multi-scale convolution module;
module M4: extracting signal dependent characteristics from the signal processed by the module M3 by using LSTM;
module M5: constructing a diesel engine fire diagnosis model by using a Keras packet under a TensorFlow framework, training, and evaluating the performance of a module according to the diagnosis result and the actual comparison;
the signal after noise reduction of the module M2 is the original signal minus the output signal of the residual convolution preprocessing module;
the loss function in the module M2 is:
wherein loss is β For the constructed loss function, n is the number of samples transferred to the model per iteration, β is the preset noise template, output i Outputting a noise signal obtained by the ith sample signal after the residual convolution preprocessing module, and outputting beta i Representing the random noise signal carried by the i-th sample signal.
5. The diesel anti-noise and cross-noise domain misfire diagnostic system of claim 4 wherein: the multi-scale convolution module includes a plurality of convolution layer modules, each convolution layer module including different branches and different sizes of convolution kernels.
6. The diesel anti-noise and cross-noise domain misfire diagnostic system of claim 5 wherein: and the convolution layer modules are mutually connected in series to extract fault characteristics of different time scales.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210589677.8A CN114858467B (en) | 2022-05-26 | 2022-05-26 | Diesel engine anti-noise and cross-noise domain fire diagnosis method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210589677.8A CN114858467B (en) | 2022-05-26 | 2022-05-26 | Diesel engine anti-noise and cross-noise domain fire diagnosis method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114858467A CN114858467A (en) | 2022-08-05 |
| CN114858467B true CN114858467B (en) | 2023-05-26 |
Family
ID=82641461
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210589677.8A Active CN114858467B (en) | 2022-05-26 | 2022-05-26 | Diesel engine anti-noise and cross-noise domain fire diagnosis method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114858467B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109784242A (en) * | 2018-12-31 | 2019-05-21 | 陕西师范大学 | EEG Noise Cancellation based on one-dimensional residual error convolutional neural networks |
| CN111329445A (en) * | 2020-02-20 | 2020-06-26 | 广东工业大学 | Atrial fibrillation identification method based on group convolution residual error network and long-term and short-term memory network |
| WO2021147095A1 (en) * | 2020-01-23 | 2021-07-29 | Baidu.Com Times Technology (Beijing) Co., Ltd. | Deep residual network for color filter array image denoising |
| CN113435322A (en) * | 2021-06-25 | 2021-09-24 | 西安交通大学 | Main shaft bearing fault detection method, system, equipment and readable storage medium |
| CN113673346A (en) * | 2021-07-20 | 2021-11-19 | 中国矿业大学 | Motor vibration data processing and state recognition method based on multi-scale SE-Resnet |
| WO2022083026A1 (en) * | 2020-10-21 | 2022-04-28 | 华中科技大学 | Ultrasound image denoising model establishing method and ultrasound image denoising method |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110595775B (en) * | 2019-09-19 | 2021-03-30 | 电子科技大学 | Rolling bearing fault diagnosis method based on multi-branch multi-scale convolutional neural network |
| CN111651937B (en) * | 2020-06-03 | 2023-07-25 | 苏州大学 | Method for diagnosing faults of in-class self-adaptive bearing under variable working conditions |
| AU2020103681A4 (en) * | 2020-11-26 | 2021-02-04 | Anhui University Of Technology | Rolling Bearing Fault Diagnosis Method Based on Fourier Decomposition and Multi-scale Arrangement Entropy Partial Mean Value |
| CN113052218A (en) * | 2021-03-15 | 2021-06-29 | 杭州电子科技大学 | Multi-scale residual convolution and LSTM fusion performance evaluation method for industrial process |
-
2022
- 2022-05-26 CN CN202210589677.8A patent/CN114858467B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109784242A (en) * | 2018-12-31 | 2019-05-21 | 陕西师范大学 | EEG Noise Cancellation based on one-dimensional residual error convolutional neural networks |
| WO2021147095A1 (en) * | 2020-01-23 | 2021-07-29 | Baidu.Com Times Technology (Beijing) Co., Ltd. | Deep residual network for color filter array image denoising |
| CN111329445A (en) * | 2020-02-20 | 2020-06-26 | 广东工业大学 | Atrial fibrillation identification method based on group convolution residual error network and long-term and short-term memory network |
| WO2022083026A1 (en) * | 2020-10-21 | 2022-04-28 | 华中科技大学 | Ultrasound image denoising model establishing method and ultrasound image denoising method |
| CN113435322A (en) * | 2021-06-25 | 2021-09-24 | 西安交通大学 | Main shaft bearing fault detection method, system, equipment and readable storage medium |
| CN113673346A (en) * | 2021-07-20 | 2021-11-19 | 中国矿业大学 | Motor vibration data processing and state recognition method based on multi-scale SE-Resnet |
Non-Patent Citations (3)
| Title |
|---|
| 基于单振动传感器BP神经网络的柴油机失火诊断;贾继德;贾翔宇;韩佳佳;任刚;;军事交通学院学报(11);第31-34页 * |
| 基于多通道一维卷积神经网络特征学习的齿轮箱故障诊断方法;叶壮;余建波;;振动与冲击(第20期);第55-65页 * |
| 强噪声背景下的柴油机失火故障诊断;刘鑫;贾云献;张英波;张艳明;;车用发动机(04);第16-20页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114858467A (en) | 2022-08-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Capriglione et al. | On-line sensor fault detection, isolation, and accommodation in automotive engines | |
| US6240343B1 (en) | Apparatus and method for diagnosing an engine using computer based models in combination with a neural network | |
| US8131484B2 (en) | Method for preprocessing vibro-sensor signals for engine diagnostics and device for carrying out thereof | |
| Nyberg et al. | Model based diagnosis for the air intake system of the SI-engine | |
| CN110147849B (en) | Method for identifying weak link of diesel engine electric control system under strong electromagnetic pulse | |
| Tang et al. | Adaptive recursive variational mode decomposition for multiple engine faults detection | |
| CN114995346A (en) | Engine electrical diagnosis calibration method, device, equipment and medium | |
| CN114858467B (en) | Diesel engine anti-noise and cross-noise domain fire diagnosis method and system | |
| CN116662920B (en) | Abnormal data identification method, system, equipment and medium for drilling and blasting method construction equipment | |
| CN117877256A (en) | Vehicle fault prediction method and system based on digital twin | |
| CN117591924A (en) | Bearing fault detection method and device, storage medium and electronic equipment | |
| CN116340848B (en) | Engine fault diagnosis method based on digital twin | |
| JP4235116B2 (en) | Apparatus and method for testing transient characteristics of internal combustion engine | |
| CN114962031B (en) | Method and system for detecting coking of pipeline of air inlet system of internal combustion engine and vehicle | |
| Watzenig et al. | Engine state monitoring and fault diagnosis of large marine diesel engines. | |
| CN114622990B (en) | Engine fire risk identification method, device, equipment and medium | |
| CN116541771A (en) | Unbalanced sample bearing fault diagnosis method based on multi-scale feature fusion | |
| CN115467752B (en) | Method, system and computer storage medium for diagnosing and analyzing fire of automobile engine | |
| CN116577104A (en) | Engine fault detection method and equipment | |
| CN114417941A (en) | Crankshaft fault detection method and system | |
| CN114898174A (en) | Cable fault recognition system based on different recognition models | |
| CN112560252A (en) | Prediction method for residual life of aircraft engine | |
| Antory | Fault diagnosis application in an automotive diesel engine using auto-associative neural networks | |
| CN115217625B (en) | Engine misfire diagnosis calibration method, device, storage medium, equipment and system | |
| KR101387421B1 (en) | Vehicle Analysis Apparatus Using Earth Level Signal |
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 |