CN109359524B - Loader condition identification model construction and identification method - Google Patents
Loader condition identification model construction and identification method Download PDFInfo
- Publication number
- CN109359524B CN109359524B CN201811043486.1A CN201811043486A CN109359524B CN 109359524 B CN109359524 B CN 109359524B CN 201811043486 A CN201811043486 A CN 201811043486A CN 109359524 B CN109359524 B CN 109359524B
- Authority
- CN
- China
- Prior art keywords
- identification
- signal data
- loader
- identification signal
- feature
- 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
- 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
- G06F2218/06—Denoising by applying a scale-space analysis, e.g. using wavelet analysis
-
- E—FIXED CONSTRUCTIONS
- E02—HYDRAULIC ENGINEERING; FOUNDATIONS; SOIL SHIFTING
- E02F—DREDGING; SOIL-SHIFTING
- E02F9/00—Component parts of dredgers or soil-shifting machines, not restricted to one of the kinds covered by groups E02F3/00 - E02F7/00
- E02F9/26—Indicating devices
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Signal Processing (AREA)
- Mining & Mineral Resources (AREA)
- Civil Engineering (AREA)
- Structural Engineering (AREA)
- Operation Control Of Excavators (AREA)
Abstract
The invention discloses a loader condition identification model construction and identification method, which comprises the steps of firstly arranging corresponding sensors on a loader for acquiring multi-source signals such as torque, pressure, gears, braking and the like, normalizing data, stripping zero drift signals, interpolating and supplementing missing values, and performing noise reduction and filtering processing on the acquired signals; secondly, selecting characteristic attributes with high contribution degree from the multi-attribute data of the loader by using a principal component analysis method, and extracting the characteristics of the principal components by using a statistical analysis method; secondly, establishing a loader working condition sample, establishing association mapping between a load signal and a pre-classification working condition mode by adopting a data mining algorithm with supervised learning, and training a large number of data samples to form a working condition identification model; the feature extraction method of principal component analysis is combined with the KNN algorithm, and a distance formula in the KNN algorithm is improved, so that the distance formula is more suitable for working condition recognition, and the accuracy and efficiency of the working condition recognition algorithm are improved.
Description
Technical Field
The invention relates to a working condition identification method, in particular to a loader working condition identification model construction and identification method.
Background
Along with the economic development of China, the yield, sales volume and holding capacity of engineering vehicles are rapidly improved, and the engineering machinery develops abnormally and rapidly. Hydraulic transmission is adopted to the product of engineering machine more than 95%, so that obtain big moment of torsion, satisfy the requirement of big inertia load, because the operational environment is abominable, the operation operating mode is complicated changeable and equipment is automatic, the informationization degree improves constantly, how to ensure that engineering machine is reliable, the high-efficient operation, be the technological problem who urgently waits to solve at present, in order to solve these problems, just need carry out the analysis to the load spectrum of loader operation, including the extraction of signal characteristic, the division of operation stage and the discernment of operation operating mode, wherein the operating mode of loader includes: no-load forward, spading, full-load backward, full-load forward, unloading and no-load backward.
The existing working condition identification method mainly judges the behavior exceeding a variation threshold value by detecting the variation of a big cavity pressure signal of a rotating bucket and a big cavity pressure signal of a movable arm during the operation of a loader and setting the variation threshold value.
Disclosure of Invention
The invention aims to provide a loader working condition identification model building and identifying method, which is used for solving the problems that the loader working condition identification method in the prior art is low in identification accuracy and the like.
In order to realize the task, the invention adopts the following technical scheme:
a loader condition identification model construction method comprises the following steps:
step 1, collecting a plurality of groups of identification signal data of a loader under different working conditions to obtain an identification signal data set; each group of identification signal data in the identification signal data set corresponds to a working condition tag to obtain an identification working condition tag set;
the operating condition label comprises: no-load forward, spading, full-load backward, full-load forward, unloading and no-load backward;
step 2, preprocessing each group of identification signal data in the identification signal data set to obtain a preprocessed identification signal data set;
the pretreatment specifically comprises the following steps:
step 21, normalizing the identification signal data set to be between 0 and 1 to obtain a second identification signal data set;
step 22, after the second identification signal data set is subjected to stripping trend item, abnormal data is removed by using a 3 sigma method, and then a null item is inserted and supplemented by using a Newton interpolation method, so as to obtain a third identification signal data set;
step 23, filtering the third identification signal data set by adopting a wavelet packet denoising method to obtain a preprocessed identification signal data set;
in the wavelet packet denoising method, a db9-6 wavelet basis is selected as a wavelet basis;
step 3, processing the preprocessed identification signal data set by adopting a dimension reduction feature extraction method to obtain an identification feature set;
the identification feature set comprises a plurality of feature samples, the number of the feature samples is the same as the group number of the identification signal data acquired in the step 1, each feature sample comprises I identification feature quantities, and I is a positive integer;
obtaining a contribution rate of the feature recognition amount: the ith identification characteristic quantity of the p characteristic sample in the plurality of characteristic samples
The ith identification feature quantity of the qth feature sample
The contribution rates of the identification characteristic quantities are the same, namely the contribution rates of I identification characteristic quantities are obtained, I belongs to [1, I ∈]P and q are positive integers, and p is not equal to q;
step 4, taking the identification feature set as input, taking the identification working condition tag set as output, training a KNN model, and obtaining a loader working condition identification model, wherein in the KNN model, the distance Dis between a p-th characteristic sample and a q-th characteristic sample is as follows: .
Wherein, CiContribution ratio for i-th identification feature quantity。
Further, in the step 4, the value of the K value in the KNN model is any positive integer within 5.
Further, in the step 3, the preprocessed identification signal data set is processed by adopting a principal component analysis method to obtain an identification signal feature set.
A loader condition identification method, the method comprising: the method adopts the working condition identification model described in any one of claims 1 to 3 to identify the signal data to be identified of the loader processed by the steps 1 to 3 described in any one of claims 1 to 3.
Compared with the prior art, the invention has the following technical characteristics:
1. the preprocessing method of the signal data provided by the invention is more accurate, so that the identification accuracy is improved, and the method comprises normalization, stripping trend items, interpolation missing values, filtering processing and the like, a wavelet packet optimal basis decomposition tree filtering method is adopted, a large number of interference signals are removed, high-frequency details are kept, a large number of signal characteristics can be kept for working condition identification, and compared with the existing technologies, such as a Butterworth filtering method, a Fourier transform filtering method, a wavelet transform filtering method and the like;
2. the accuracy is low if only single pressure or torque is adopted as a judgment variable for identification in the prior art; if all the collected data are identified, the operation amount is too large, the calculation time is too long, the delay is serious, and the condition identification is not facilitated;
3. the method for extracting the dimensionality reduction features of the principal component analysis method is combined with the KNN algorithm, and a distance formula in the KNN algorithm is improved, so that the distance formula is more consistent with working condition identification, and the accuracy and efficiency of the working condition identification algorithm are improved.
Drawings
FIG. 1 is a flow chart of a method for constructing a loader condition identification model provided by the invention.
Detailed Description
The following are specific examples provided by the inventors to further explain the technical solutions of the present invention.
Example one
The embodiment discloses a method for constructing a working condition identification model of a loader, which comprises the following steps:
step 1, collecting a plurality of groups of identification signal data of a loader under different working conditions as an identification signal data set; each group of identification signal data in the identification signal data set corresponds to a working condition tag to obtain an identification working condition tag set;
the working condition of the loader refers to a working state of the loader under a condition directly related to the action thereof, and generally, the working condition of the loader includes shoveling, full-load transportation, and unloading.
In the embodiment, the working conditions of the loader are finely divided to ensure the accuracy of judgment, and the working conditions of the loader comprise no-load forward, excavation, full-load backward, full-load forward, unloading and no-load backward;
the identification signal data of the loader under different working conditions comprises: the torque of a front axle of the loader, the rotating speed of the front axle of the loader, the torque of a rear axle of the loader, the pressure of a working pump, the pressure of a steering pump, the flow of the working pump, the flow of the steering pump, the rotating speed of an engine, a braking signal, an accelerator signal, a gearbox signal and the like.
In the multiple signals, the subjectivity of the braking signal, the throttle signal, the flow of the working pump, the flow of the steering pump and the gearbox signal is too strong, and the subjectivity is too strong because the throttle opening, the braking signal and the gearbox signal belong to human operation factors in theoretical analysis; the flow rates of the working pump and the steering pump only can reflect the working speed of the hydraulic oil cylinder, and are also greatly influenced by human factors. When the loader condition is intelligently identified, once the behavior signal of the driver is added, although the working condition can be obviously judged, the driver is required to operate without error, otherwise, the error identification is easy, and therefore, the range of the identification signal is not included.
The operating condition label comprises: no-load forward, spading, full-load backward, full-load forward, unloading and no-load backward;
according to the collected driver behavior data, after preprocessing, the working condition segments are divided, and the working condition segments are shown in table 1.
TABLE 1 judgment basis for loader behavior
The criteria in table 1 are universal. Judging whether the vehicle is advancing in an unloaded mode at S1 or in a full-loaded mode at S4 according to the gear; the stage S2 is that braking exists at the forward gear, and is followed by a large accelerator opening, so that the preparation of the material pile for digging is near, if no braking exists, the impact of the vehicle is too large, and the large accelerator opening is the work during digging; at the end of the stage S2, the vehicle speed is almost 0 due to excessive load, and sometimes a slight braking signal is accompanied, so the reverse gear signal is the basis for the determination of the start of the stage S3; the stage S5 is an unloading stage, the accelerator is usually loosened and the brake is usually stepped to avoid collision when the vehicle approaches a dump truck, and the braking time is slightly long at the moment because the unloading and gear shifting are required; the stage S6 is after the brake signal is finished, before the throttle signal, but not after the shift signal, because there are cases where the shift is finished and the unloading is not finished yet.
Therefore, in this embodiment, the identification signal data set includes { loader front axle torque, loader front axle rotational speed, loader rear axle torque, working pump pressure, steering pump pressure, and engine rotational speed }, and the values of the above 6 identification signal data are collected under 6 different conditions of the loader, that is, each condition corresponds to one set of identification signal data, and one set of identification signal data includes 6 values. The operating condition label is the name of the operating condition, and in the embodiment, the label comprises 1-no-load forward, 2-shovel, 3-full-load backward, 4-full-load forward, 5-unload and 6-no-load backward.
Step 2, preprocessing each group of identification signal data in the identification signal data set to obtain a preprocessed identification signal data set; all identification signal data in the preprocessed identification data set are between 0 and 1;
in order to improve the running speed of the algorithm and ensure the identification result, the identification signal data is preprocessed, which comprises the following steps:
step 21, normalizing the identification signal data set to be between 0 and 1 to obtain a second identification signal data set;
in the present embodiment, in order to improve the efficiency of the algorithm, the units of identification signal data are converted into corresponding dimensionless numbers. Taking the engine speed as an example, the engine speed is subjected to maximum-maximum normalization, and the value of the engine speed is mapped into the [0,1] interval.
Step 22, after the stripping trend item is carried out on the second identification signal data set, removing abnormal data in the second identification signal data set, and then inserting and supplementing a blank item to obtain a third identification signal data set;
in this step, since the identification signals are all collected by the sensors, the environmental disturbance is amplified. The acquired vibration signal tends to deviate from the baseline in the time series, and a linear general trend (zero drift) is generated, and the process of the trend changing along with the time is called a trend term. The period of the trend term is far larger than the frequency of the sample, which causes great distortion in correlation analysis, power spectrum analysis and even signal distortion. Therefore, when analysis is performed after a signal is measured for a long time, it is necessary to strip a trend term from data.
As a preferred implementation mode, the algorithm adopting the least square method stripping trend term is simple and high in precision, and not only can eliminate the approximately linear growth trend, but also can eliminate the trend of a high-order polynomial.
In the step, after the identification signal passes through the stripping trend term, an abnormal signal still exists, and the identification signal is characterized by strong randomness, large amplitude, undefined period and the like. If the data containing abnormal values are calculated without being eliminated, the result of calculation analysis is influenced.
Commonly used interpolation methods include lagrange interpolation, newton interpolation, and the like.
As a preferred embodiment, 3 σ principle is adopted to eliminate abnormal data in the identification signal data set with the stripping trend term, and then a null term is inserted and supplemented by using a newton interpolation method to obtain a third identification signal data set.
Since under a large number of sample data, normal distribution is usually followed, while the 3 σ rule defines values that deviate more than three times the standard deviation from the mean among the measured data as outliers. Under the 3 sigma principle, the probability of data anomaly is P (| x-mu | >3 sigma) ≦ 0.003, belonging to the least probable event, while the interval of the overall main distribution is (mu-3 sigma, mu +3 sigma).
In the data preprocessing stage, abnormal data should be removed, but neglecting missing values and removed abnormal values discards a large amount of information hidden in records, which causes waste of acquired information, so that missing values and removed abnormal points need to be interpolated, when interpolation nodes need to be added, the basis function of lagrange interpolation changes accordingly, which is inconvenient in calculation and practice, and thus, a newton interpolation method is selected.
And step 23, filtering the third identification signal data set to obtain a preprocessed identification signal data set.
Because the interference of signals such as noise causes random errors in the acquired data, and the fluctuation of the errors affects the true value, the noise signals need to be filtered, so in this step, the third identification signal data set is filtered, and the commonly used filtering methods include fast fourier transform, butterworth filtering, wavelet analysis, wavelet packet denoising, and the like.
In this embodiment, the filtering effect is judged by comparing the superiority and inferiority of the filtering processing method with a large number of experiments and by using the signal-to-noise ratio, the root-mean-square error and the peak error, and the result is shown in table 2.
TABLE 2 evaluation of denoising Effect
As can be seen from Table 2, the signal-to-noise ratio, the root mean square error and the peak error of the wavelet analysis and the wavelet packet analysis are similar and superior to the Butterworth denoising. Although the evaluation result of the denoising method of the wavelet transform is slightly better than the denoising of the wavelet packet, the decomposition of the wavelet packet is finer, part of high-frequency details can be reserved, and the operation is simpler and more convenient, so that as a preferred implementation mode, the wavelet packet denoising method is adopted for filtering the third identification signal data set to obtain a preprocessed identification signal data set;
in the wavelet packet denoising method, the wavelet basis is selected from db9-6 wavelet basis.
Through the processing of the steps 21 to 23, six kinds of signal identification data, namely, the torque of the front axle of the loader, the rotating speed of the front axle of the loader, the torque of the rear axle of the loader, the pressure of a working pump, the pressure of a steering pump and the rotating speed of an engine, are preprocessed to be in a range of 0-1, and the stationarity of the identification signal data is ensured.
However, in this embodiment, the parameters such as the engine power, the transmission system output power, the steering pump and the operating pump power, which are calculated based on the 6 kinds of signal identification data, i.e., the loader front axle torque, the loader front axle rotational speed, the loader rear axle torque, the operating pump pressure, the steering pump pressure, and the engine rotational speed, have a certain correlation with the original variables, but can more intuitively reflect the operating state, belong to the derivative variables, and are also used as the identification signals in this embodiment.
Through the steps, each group of preprocessed identification data in the preprocessed identification data set comprises 11 identification signals of loader front axle torque, loader front axle rotating speed, loader rear axle torque, pressure of a working pump, steering pump pressure, power of the working pump, power of the steering pump, engine rotating speed, engine torque, output axle power and engine power.
Step 3, processing the preprocessed identification signal data set by adopting a dimension reduction feature extraction method to obtain an identification feature set;
the identification feature set comprises a plurality of feature samples, the number of the feature samples is the same as the group number of the identification signal data acquired in the step 1, each feature sample comprises I identification feature quantities, and I is a positive integer;
obtaining a contribution rate of the feature recognition amount: the ith identification characteristic quantity of the p characteristic sample in the plurality of characteristic samples
The ith identification feature quantity of the qth feature sample
The contribution rates of the identification characteristic quantities are the same, namely the contribution rates of I identification characteristic quantities are obtained, I belongs to [1, I ∈]P and q are positive integers, and p is not equal to q;
because the identification signal data after preprocessing is also high-dimensional data which contains attributes which are not strong or even irrelevant to the relation degree of the judgment mode, dimension disaster can be caused if sample coefficients, Euclidean distances and the like are calculated under high-dimensional numbers, and the operation amount is relatively large if the identification signals are all input into the identification model, so that part of main signals need to be selected from the identification signal data, the operation amount is reduced, the accuracy is improved as much as possible, and an important way for relieving the dimension disaster is to screen important features from the attributes. Although domain experts can sort out useful attributes, ignoring some of the relevant attributes or retaining irrelevant attributes can result in a degradation of the quality of the operating condition intelligent algorithm.
The existing dimension reduction feature extraction methods comprise a principal component analysis method, an LBP feature extraction method and the like.
In a preferred embodiment, the preprocessed identification signal data set is processed by principal component analysis to obtain an identification signal feature set.
In this embodiment, feature extraction is performed on 11 input identification signals by a principal component analysis method, and when the dimension reduction is set to 3, the 11 identification signals can be reduced to 3 identification feature quantities, namely, front axle torque, rear axle torque, and main pump power.
When the PCA method is used to perform the dimension reduction feature extraction, the contribution rate of each feature recognition amount can be obtained accordingly, and in this embodiment, the contribution rate of each feature recognition amount is shown in table 3.
TABLE 3 principal component analysis contribution rate
Each feature sample includes the same number and kind of identification feature quantities, in this embodiment, each feature sample includes 3 identification feature quantities, which are front axle torque, rear axle torque and main pump power, respectively, and each feature sample corresponds to one operating condition label, that is, includes 3 identification feature quantities under 6 operating condition labels, for example: the characteristic sample 1 is [ front-rear torque, rear axle torque, main pump power ] ([ 0.362,0.861,0.153 ]), and the working condition corresponding to the characteristic sample is 3-full load backward; the characteristic sample 2 is [ front-rear torque, rear axle torque, main pump power ] ═ 0.421,0.937,0.268], and the working condition corresponding to the characteristic sample is 1-no-load forward.
Step 4, taking the identification feature set as input, taking the identification working condition tag set as output, training a KNN model, and obtaining a loader working condition identification model, wherein in the KNN model, the distance Dis between a p-th characteristic sample and a q-th characteristic sample is as follows:
wherein p and q are both positive integers,
i is more than or equal to 1 and less than or equal to I which is the ith identification characteristic quantity of the p identification sample, I is the total number of the identification characteristic quantities in the characteristic sample, I>0，
Identifying ith of sample for qthSpecies identification feature quantity, CiThe contribution ratio of the identification feature quantity of the i-th type.
In the present embodiment, for the 1 st identification sample [ front-rear torque, rear axle torque, main pump power ] (0.5, 0.3, 0.2) and the 2 nd identification sample [ front-rear torque, rear axle torque, main pump power ] (0.6, 0.4, 0.1) in the identification feature set, the distance Dis between them is:
in the embodiment, the KNN recognition algorithm and the PCA dimension reduction method are fused, so that the accuracy of the recognition algorithm is improved, and the algorithm recognition efficiency is improved.
In this embodiment, another key point of the KNN recognition algorithm is to determine a K value, and if the K value is smaller, the model becomes complex and is more sensitive to neighboring training points, and overfitting is likely to occur; when the K value is large, the model is too simple, and the training points with longer distance also play a role, so that the fitting is easy to be underfitted.
In a preferred embodiment, K ═ 1,5, and the value of K is an integer between 1 and 5.
The KNN algorithm is only related to a small amount of adjacent cost when working condition mode intelligent identification is carried out, the problem of uneven distribution among samples can be avoided, judgment is carried out by means of limited adjacent samples, and the method is more suitable for sample sets with more cross overlapping. But the defect is large calculation amount, so the invention adopts the principal component analysis method to extract the principal component with the highest contribution rate and eliminates the attribute with small effect so as to achieve the purpose of reducing the storage amount and the calculation amount.
The KNN algorithm has good accuracy in the aspect of low dimensionality, but the calculation is complex in high dimensionality and higher memory is occupied, so that principal component analysis is performed before the KNN model is obtained, main attributes in a data set are extracted, and the operation amount is reduced.
Example two
The invention also discloses a loader condition identification method, which comprises the following steps:
and identifying the signal data to be identified of the loader processed by the steps 1 to 3 in the first embodiment by adopting the working condition identification model in the first embodiment.
In this embodiment, the data of the signal to be recognized is [ loader front axle torque, loader front axle rotation speed, loader rear axle torque, [1450,1280,2870,8.8,16.9,2430], after being processed in steps 1 to 3 of the first embodiment, the feature set of the recognition obtained is [ front axle torque, rear axle torque, main pump power ] (0.451, 0.942,0.287], and after being recognized by the behavior recognition model, the recognition result is 2-shovel.
Claims (3)
1. A loader condition identification model construction method comprises the following steps:
step 1, collecting a plurality of groups of identification signal data of a loader under different working conditions to obtain an identification signal data set; each group of identification signal data in the identification signal data set corresponds to a working condition tag to obtain an identification working condition tag set;
the method is characterized in that:
the operating condition label comprises: no-load forward, spading, full-load backward, full-load forward, unloading and no-load backward;
step 2, preprocessing each group of identification signal data in the identification signal data set to obtain a preprocessed identification signal data set;
the pretreatment specifically comprises the following steps:
step 21, normalizing the identification signal data set to be between 0 and 1 to obtain a second identification signal data set;
step 22, after the second identification signal data set is subjected to stripping trend item, abnormal data is removed by using a 3 sigma method, and then a null item is inserted and supplemented by using a Newton interpolation method, so as to obtain a third identification signal data set;
step 23, filtering the third identification signal data set by adopting a wavelet packet denoising method to obtain a preprocessed identification signal data set;
in the wavelet packet denoising method, a db9-6 wavelet basis is selected as a wavelet basis;
step 3, processing the preprocessed identification signal data set by adopting a dimension reduction feature extraction method to obtain an identification feature set;
step 3, processing the preprocessed identification signal data set by adopting a principal component analysis method to obtain an identification signal feature set;
the identification feature set comprises a plurality of feature samples, the number of the feature samples is the same as the group number of the identification signal data acquired in the step 1, each feature sample comprises I identification feature quantities, and I is a positive integer;
obtaining a contribution rate of the feature recognition amount: the ith identification characteristic quantity of the p characteristic sample in the plurality of characteristic samples
The ith identification feature quantity of the qth feature sample
The contribution rate of (c) is the same, I is equal to [1, I ∈ [ ]]P and q are positive integers, and p is not equal to q, namely the contribution rate of I identification characteristic quantities is obtained;
step 4, taking the identification feature set as input, taking the identification working condition tag set as output, training a KNN model, and obtaining a loader working condition identification model, wherein in the KNN model, the distance Dis between a p-th characteristic sample and a q-th characteristic sample is as follows:
wherein, CiThe contribution ratio of the i-th recognition feature amount.
2. The method for constructing the loader working condition identification model according to claim 1, wherein in the step 4, the value of the K value in the KNN model is any positive integer within 5.
3. A loader condition identification method, comprising: the method adopts the working condition identification model described in any one of claims 1-2 to identify the signal data to be identified of the loader processed by the steps 2-3 described in any one of claims 1-2.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811043486.1A CN109359524B (en) | 2018-09-07 | 2018-09-07 | Loader condition identification model construction and identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811043486.1A CN109359524B (en) | 2018-09-07 | 2018-09-07 | Loader condition identification model construction and identification method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109359524A CN109359524A (en) | 2019-02-19 |
| CN109359524B true CN109359524B (en) | 2021-06-22 |
Family
ID=65350578
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811043486.1A Active CN109359524B (en) | 2018-09-07 | 2018-09-07 | Loader condition identification model construction and identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109359524B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110886338A (en) * | 2019-12-19 | 2020-03-17 | 天津工程机械研究院有限公司 | Semi-automatic shovel loading control system and method for loader |
| CN110847277A (en) * | 2019-12-19 | 2020-02-28 | 天津工程机械研究院有限公司 | Intelligent switching system and control method for power mode of loader engine |
| CN112700085A (en) * | 2020-12-11 | 2021-04-23 | 华南理工大学 | Association rule based method, system and medium for optimizing steady-state operation parameters of complex system |
| CN113502870B (en) * | 2021-07-15 | 2023-02-07 | 上海三一重机股份有限公司 | Excavator working condition judging method and device |
| CN113565162B (en) * | 2021-07-27 | 2022-09-23 | 徐州徐工挖掘机械有限公司 | Excavator, accurate control method and system thereof, vehicle-mounted end and storage medium |
| CN113882457B (en) * | 2021-09-27 | 2023-02-14 | 徐州徐工挖掘机械有限公司 | Hydraulic excavator working condition identification and statistics method and system and hydraulic excavator |
| CN113954843B (en) * | 2021-11-12 | 2023-06-27 | 燕山大学 | Real-time working condition identification method for hydraulic mechanical stepless speed change loader |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103839057A (en) * | 2014-03-28 | 2014-06-04 | 中南大学 | Antimony floatation working condition recognition method and system |
| CN104794492A (en) * | 2015-04-27 | 2015-07-22 | 重庆大学 | Online machine tool equipment machining and running state recognizing method based on power feature models |
| CN106055778A (en) * | 2016-05-27 | 2016-10-26 | 大连楼兰科技股份有限公司 | Remote damage-assessment system and method established based on artificial intelligence for different types of vehicles |
| CN106096593A (en) * | 2016-07-22 | 2016-11-09 | 厦门大学 | A kind of recognition methods of the effective operation section loading mechanical shovel process of assembling |
| CN106778533A (en) * | 2016-11-28 | 2017-05-31 | 国网上海市电力公司 | PCA KSICA energy-storage system typical condition recognition methods based on kernel function |
| CN107908874A (en) * | 2017-11-15 | 2018-04-13 | 上海华兴数字科技有限公司 | Operating mode's switch method and apparatus and engineering mechanical device |
| CN108052975A (en) * | 2017-12-12 | 2018-05-18 | 浙江大学宁波理工学院 | It is a kind of that real-time working condition Forecasting Methodology is run based on the vehicle of core pivot and neutral net |
| CN108280543A (en) * | 2018-01-17 | 2018-07-13 | 杭州安脉盛智能技术有限公司 | The adaptive equipment health degree appraisal procedure of operating mode based on classification Regression Model |
-
2018
- 2018-09-07 CN CN201811043486.1A patent/CN109359524B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103839057A (en) * | 2014-03-28 | 2014-06-04 | 中南大学 | Antimony floatation working condition recognition method and system |
| CN104794492A (en) * | 2015-04-27 | 2015-07-22 | 重庆大学 | Online machine tool equipment machining and running state recognizing method based on power feature models |
| CN106055778A (en) * | 2016-05-27 | 2016-10-26 | 大连楼兰科技股份有限公司 | Remote damage-assessment system and method established based on artificial intelligence for different types of vehicles |
| CN106096593A (en) * | 2016-07-22 | 2016-11-09 | 厦门大学 | A kind of recognition methods of the effective operation section loading mechanical shovel process of assembling |
| CN106778533A (en) * | 2016-11-28 | 2017-05-31 | 国网上海市电力公司 | PCA KSICA energy-storage system typical condition recognition methods based on kernel function |
| CN107908874A (en) * | 2017-11-15 | 2018-04-13 | 上海华兴数字科技有限公司 | Operating mode's switch method and apparatus and engineering mechanical device |
| CN108052975A (en) * | 2017-12-12 | 2018-05-18 | 浙江大学宁波理工学院 | It is a kind of that real-time working condition Forecasting Methodology is run based on the vehicle of core pivot and neutral net |
| CN108280543A (en) * | 2018-01-17 | 2018-07-13 | 杭州安脉盛智能技术有限公司 | The adaptive equipment health degree appraisal procedure of operating mode based on classification Regression Model |
Non-Patent Citations (2)
| Title |
|---|
| "An Optical Surface Inspection and Automatic Classification Technique Using the Rotated Wavelet Transform";Raunak Borwankar等;《IEEE Transactions on Instrumentation and Measurement》;20180105;690-697页 * |
| "液力变矩器与发动机的功率匹配与转速感应控制";张泽宇;《机械科学与技术》;20180131;第37卷(第1期);55-62页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109359524A (en) | 2019-02-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109359524B (en) | Loader condition identification model construction and identification method | |
| CN108038553B (en) | Rolling mill equipment state on-line monitoring and diagnosing system and monitoring and diagnosing method | |
| CN111562108A (en) | Rolling bearing intelligent fault diagnosis method based on CNN and FCMC | |
| Sun et al. | Multiclassification of tool wear with support vector machine by manufacturing loss consideration | |
| CN111539374B (en) | Rail train bearing fault diagnosis method based on multidimensional data space | |
| CN109237007B (en) | Efficiency optimization control method for loader transmission system | |
| CN112380992B (en) | Method and device for evaluating and optimizing accuracy of monitoring data in machining process | |
| CN112801555B (en) | Vehicle dynamic property comprehensive evaluation method based on Internet of vehicles big data | |
| CN114662954A (en) | Vehicle performance evaluation system | |
| CN106446081A (en) | Method for mining association relationship of time series data based on change consistency | |
| CN111811819A (en) | Bearing fault diagnosis method and device based on machine learning | |
| CN115220396A (en) | Intelligent monitoring method and system for numerical control machine tool | |
| CN112881006B (en) | Gear fault diagnosis method | |
| CN113987927A (en) | Commercial vehicle-mounted quality detection method | |
| CN101826070A (en) | Key point-based data sequence linear fitting method | |
| CN110847907A (en) | Coal rock identification method based on touch sense | |
| CN113029619A (en) | Underground scraper fault diagnosis method based on C4.5 decision tree algorithm | |
| CN112270209A (en) | Method for extracting vibration signal characteristics of threshing cylinder of combine harvester | |
| CN112885080B (en) | Construction method for driving condition of new energy automobile | |
| CN112528753A (en) | Preprocessing method for impact vibration signals in rolling process of rolling mill | |
| Zimroz et al. | Novel techniques of diagnostic data processing for belt conveyor maintenance | |
| CN117807443B (en) | Training method of tractor working condition identification model and tractor working condition identification method | |
| CN115971970B (en) | Milling cutter abrasion monitoring method based on multi-parameter guiding spatial attention mechanism | |
| CN110097133A (en) | A kind of preprocess method for shield machine big data | |
| CN113433925B (en) | Predictive control method and device for heading machine motor 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |