CN112099477A - Fault tracing method in lithium ion battery production process - Google Patents
Fault tracing method in lithium ion battery production process Download PDFInfo
- Publication number
- CN112099477A CN112099477A CN202011219367.4A CN202011219367A CN112099477A CN 112099477 A CN112099477 A CN 112099477A CN 202011219367 A CN202011219367 A CN 202011219367A CN 112099477 A CN112099477 A CN 112099477A
- Authority
- CN
- China
- Prior art keywords
- binary
- fault
- normal state
- binarization
- physical quantity
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0218—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults
- G05B23/0243—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults model based detection method, e.g. first-principles knowledge model
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/20—Pc systems
- G05B2219/24—Pc safety
- G05B2219/24065—Real time diagnostics
Abstract
The invention discloses a fault tracing method in the production process of a lithium ion battery, which comprises the steps of firstly obtaining a normal state sample set and a fault state sample set in the production process of the lithium ion battery, carrying out data binarization operation, constructing a binary feature combination for distinguishing normal state data from fault state data to optimize a model, solving a binary feature group set which enables a model objective function to reach an extreme value, reversely converting the binary feature group set into a generation type rule set based on physical quantity, and carrying out fault tracing analysis according to the generation type rule set. The fault tracing method does not depend on a system mechanism and prior knowledge in the production process of the lithium ion battery, a binary characteristic combination optimization model is creatively constructed from the perspective of data, and multiple fault judgment rules are discovered in parallel to realize fault tracing analysis.
Description
Technical Field
The invention relates to the field of fault diagnosis in an industrial process, in particular to a fault tracing method in a lithium ion battery production process.
Background
The lithium ion battery production process has high automation level, the integration level and the complexity are increasingly improved, and different process variables are coupled in a correlated manner, so that any tiny problems such as artificial misoperation, abnormal equipment parts and the like can cause chain reaction, the operation fault of the whole system is caused, and huge economic loss is caused. The fault detection and diagnosis technology is one of the key technologies for ensuring industrial production safety and reducing maintenance cost, and is mainly divided into a qualitative method and a quantitative method. The qualitative method analyzes and diagnoses the fault based on the experience knowledge accumulated by field experts or professional technicians, and has the advantages of clear diagnosis mechanism, strong interpretability, low fault identification rate and classification rate; quantitative methods further include model and data based methods. The method comprises the following steps of establishing a mechanism model of an industrial process based on a model method, obtaining estimated values of an intermediate variable and an output variable in a normal state according to system input simulation, and comparing measured data to identify faults and anomalies; the data-based method does not depend on system structure and parameters, and only judges whether a fault exists by constructing a data model. In contrast, due to the complexity of modern industrial systems, model-based methods (Kavuri Y S N. A view of process fault detection and diagnosis: Part I: Quantitative model-based methods [ J ]. Computers & Chemical Engineering, 2003.) have difficulty in accurately characterizing the dynamic behavior of process variables, while data-based methods (Severson K, Chaiwatanom P, Brantz R D. perspective on process monitoring of industrial system [ J ]. Annual Review in Control, 2016, 42: 190-.
However, common diagnostic models based on data have a common problem, namely the interpretability of complex models. Taking a neural network as an example, the feature transformation executed inside the neural network is a black box which is difficult to understand for users, failure cause and effect variables and influence variables cannot be obtained from the black box, and the stability of the diagnosis performance of the neural network is difficult to guarantee. The industrial system fault diagnosis method must have high reliability, and the method based on data is required to be capable of effectively searching for the change of physical quantity causing the fault, thereby revealing the cause of the fault, realizing the fault tracing analysis and improving the reliability of the diagnosis method.
Based on this, there is a need in the art for new data-based methods of fault tracing in the production of lithium ion batteries to address the above-mentioned problems.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides a fault tracing method in the production process of a lithium ion battery.
A fault tracing method in the production process of a lithium ion battery comprises the following steps:
step 10: acquiring a normal state and fault state sample set in the production process of the lithium ion battery;
step 20: performing data binarization operation on the normal state data and the fault state data in the step 10;
step 30: constructing a binary characteristic combination for distinguishing the normal state data and the fault state data after the binarization operation to optimize a model;
step 40: solving a binary feature group set which enables the model objective function in the binary feature combined optimization model constructed in the step 30 to reach an extreme value;
step 50: reversely converting the binary feature set obtained in the step 40 into a production rule set based on physical quantity;
step 60: and (4) performing fault tracing analysis according to the generated rule set obtained in the step 50.
Data before and after binarization operation can be mutually converted, the optimal combination capable of distinguishing the normal and fault state binarization characteristics is found by the optimization model, and then the optimal combination can be reversely converted into normal and fault state judgment rules corresponding to the original data.
Preferably, in step 10, the normal state and fault state samples include physical quantities capable of reflecting the production process of the lithium ion battery, and the physical quantities include process variables and control variables; and acquiring physical quantity measured values of a period of continuous time in a normal state and a fault state through a data acquisition system to form a normal state and fault state sample set.
More preferably, step 20 comprises:
step 201: mixing and sorting the normal state physical quantities and the fault state physical quantities, and searching the average value of all adjacent normal state physical quantities and fault state physical quantities as a binarization reference point;
step 202: for each physical quantity, constructing a Gaussian distribution function by using the mean value and the variance of the normal state samples, calculating the number r of the fault state samples in the range of 3 times of the standard deviation of the Gaussian distribution function of the normal state samples, and setting the upper limit and the lower limit T of a number thresholdt、TbTo obtain a binarization point;
step 203: and for each physical quantity, mapping the physical quantity into binary features according to the relative size of the value of the physical quantity and all binary point values, wherein all binary features form a binary vector.
Further preferably, in step 202, the obtaining of the binarization point includes three cases: when r > TtTaking the maximum value and the minimum value of the binarization reference points of the corresponding physical quantities as binarization points; when T isb≤r≤TtTaking the statistical value of the binary reference point corresponding to the physical quantity as a binary point; when r is less than TbThen, all the binarization reference points corresponding to the physical quantities are taken as the binarization points.
More preferably, in step 30, the binary feature combination optimization model includes three objectives and four constraints;
the three targets are: matching degree of the binarization feature group and the normal state sample, difference degree of the binarization feature group and the fault state sample and number of elements in the binarization feature group;
the four constraints include: the number of elements of the binary characteristic group is restricted, the unidirectional restriction of a production formula rule, the matching restriction of normal state samples and the difference restriction of fault state samples,
the model objective function representation is minimized to be the weighted sum of the matching degree of the binarization feature group and the normal state sample, the difference degree of the binarization feature group and the fault state sample and the number of elements in the binarization feature group.
Further preferably, the binary feature combination optimization model is expressed as:
where the first row represents the model objective function, min represents the minimization,s、m、d、lall represent the optimization variables of the system,XandYrespectively representing normal state and fault state sample sets, equations (1) - (8) representing constraints of the optimization variables,s.t.representing that the constraint conditions shown in the formulas (1) to (8) need to be satisfied;m i representing a binary feature set withiThe matching degree of the normal state samples is obtained,d i representing a binary feature set withjThe degree of difference of the samples of the individual fault states,lrepresenting the number of elements in the binarized feature set,αandβis the weight of the corresponding target;
is shown to have 2nExtended binary feature set of individual elements, frontnEach element is a binary characteristic group element, and thennIt is the element-wise inverse value of the preceding binarized feature value element,kandprespectively representing a binary feature group number and an element number in the group;
binary vector representing expansion, expansion mode and twoThe set of valued features is the same. An optimization model contains three elements: objective functions, optimization variables and constraints. The goal of the model is to search for the values of the optimization variables such that the objective function is max or min minimized, based on the constraints being satisfied.
More preferably, in step 40, the solving method of the binary feature set is a branch and bound method, a genetic algorithm or a particle swarm algorithm.
More preferably, step 50 comprises:
step 501: determining the physical quantity corresponding to the binary feature set element according to the number of the binary points corresponding to each physical quantity;
step 502: and converting the physical quantity corresponding to the binary feature group element and the value of the binary point into an if-then (if-then) generating formula rule according to the value of the binary feature group element, wherein if (if) some physical quantity > (greater than) or < (smaller than) a limit value, some fault occurs (then).
Further preferably, in step 60, the failure tracing is implemented according to the physical quantity contained in the production formula rule and the limit value corresponding to the physical quantity (the limit value is obtained by reversely converting the binary point value) to analyze the cause of the failure.
The invention aims at the problem of fault tracing in the production process of the lithium ion battery, does not depend on the system mechanism and the prior knowledge in the production process of the lithium ion battery, creatively constructs a binary characteristic combination optimization model from the perspective of data, and realizes parallel discovery of multiple fault judgment rules to realize fault tracing analysis. The invention provides a fault tracing transparency data model used in the production process of a lithium ion battery, which can be effectively expressed by a production formula rule and has good interpretability and acceptability.
Drawings
Fig. 1 is a flowchart of a method for tracing a fault in a lithium ion battery production process according to the present invention.
FIG. 2 is a diagram of the physical quantities of normal and fault samples corresponding to the Tennessee Eastman process in an embodiment of the present invention.
Fig. 3 is a schematic diagram of a lithium ion power battery slurry preparation process in an embodiment of the invention.
Detailed Description
Preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only for explaining the technical principle of the present invention, and are not intended to limit the scope of the present invention.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
The invention provides an industrial process fault tracing method based on binarization characteristic combination optimization. As shown in fig. 1, the method comprises the steps of:
step 10: acquiring a normal state and fault state sample set in an industrial process;
the normal state and fault state samples of step 10 contain the main physical quantities capable of reflecting the state of the industrial process, including process variables and control variables; and acquiring physical quantity measured values of a period of continuous time in a normal state and a fault state through a data acquisition system to form a sample set.
The normal state sample set is expressed as
The set of fault state samples is represented as
WhereinN x AndN y respectively representing the number of normal state and fault state samples,nrepresenting the sum of the quantities of the main physical quantities, including the process variables and the control variables.
In one embodiment of the invention, the Tennessee Eastman (TE) process is used as an example to obtain normal and fault state sample sets, and the data is referenced to http:// web. The TE process contains 5 units and 8 species, specifically 4 inputs A, C, D and E, 2 outputs G and H, catalyst B and by-product F. In the embodiment, 22 process variables PV (1) -PV (22) and 11 operation variables OV (1) -OV (11) are selected as main physical quantities reflecting the state of the industrial process, and 100 normal state samples and 80 fault state samples of 3 different faults F (2), F (7) and F (10) are selected to form a sample set.
Step 20: carrying out data binarization operation on normal state data and fault state data;
the step 20 specifically includes:
step 201: mixing and sorting the normal state physical quantities and the fault state physical quantities, and searching the average value of all adjacent normal state physical quantities and fault state physical quantities as a binarization reference point;
step 202: for each physical quantity, constructing a Gaussian distribution function by using the mean value and the variance of the normal state samples, calculating the number r of the fault state samples in the range of 3 times of the standard deviation of the Gaussian distribution function of the normal state samples, and setting the upper and lower limits of the number thresholdT t AndT b to obtain a binarization point;
the acquisition of the binarization points in the step 202 includes three cases: when r > TtTaking the maximum and minimum values of the corresponding physical quantity binarization reference points as binarization points; when T isb≤r≤TtConstantly, taking the statistics of the maximum value, the minimum value, the average value, the median, the mode and the like of the binary reference point of the corresponding physical quantity as the binary point; r < TbThen, all the binarization reference points corresponding to the physical quantities are taken as the binarization points.
Step 203: for each physical quantity, mapping the physical quantity into a binary characteristic according to the relative size of the value of the physical quantity and all binary point values, wherein the binary characteristic is composed of 0 and 1, and if the physical quantity is greater than the binary point values, the physical quantity is 1; otherwise it is 0. All the binarization features form a binarization vector with the length r.
Step 30: constructing a binary characteristic combination optimization model for distinguishing normal state data and fault state data;
the binary feature combination optimization model in the step 30 belongs to a constrained mixed integer linear programming problem, and comprises three targets and four main constraints; the target function representation is minimized to be the weighted sum of the matching degree of the binarization characteristic group and the normal state sample, the difference degree of the binarization characteristic group and the fault state sample and the number of elements in the binarization characteristic group; the four main constraints comprise a binarization feature set element quantity constraint (constraint 1), a production formula rule unidirectional constraint (constraint 2), a normal state sample matching constraint (constraint 3) and a fault state sample difference constraint (constraint 4).
The binary feature combinatorial optimization model is represented as:
where the first row represents the model objective function, min represents the minimization,s、m、d、lall represent the optimization variables of the system,XandYrespectively representing normal state and fault state sample sets, equations (1) - (8) representing constraints of the optimization variables,s.t.representing that the constraint conditions shown in the formulas (1) to (8) need to be satisfied;m i representing a binary feature set withiThe matching degree of the normal state samples is obtained,d i representing a binary feature set withjThe degree of difference of the samples of the individual fault states,lrepresenting the number of elements in the binarized feature set,αandβis the weight of the corresponding target;
is shown to have 2nExtended binary feature set of individual elements, frontnEach element is a binary characteristic group element, and thennThe preceding binary feature value elements are inverted by element (binary feature set is represented as a string of sequences consisting of 0 and 1, and inverting 0 and 1 by element means, for example, original binary feature set: 000111000, inverted binary feature set: 111000111),kandprespectively representing a binary feature group number and an element number in the group;
and representing an expanded binary vector, wherein the expansion mode is the same as the binary feature group.
Step 40: solving a binary feature set which enables the model objective function to reach an extreme value;
the solving method of the binary feature set in the step 40 includes a branch and bound method, a genetic algorithm, a particle swarm algorithm, and the like. In the process of solving the binary characteristic combination optimization model, when the found optimal solution can completely distinguish the normal state sample set from the fault state sample set, the calculation is finished; if part of the normal state data can not be effectively distinguished, the distinguished normal state data are deleted, and the model is solved again to obtain an additional binary feature set. All the binary feature sets constitute a set of binary feature sets.
Step 50: reversely converting the binary feature group set into a production rule set based on physical quantity;
the step 50 specifically includes:
step 501: determining the physical quantity corresponding to the binary feature set element according to the number of the binary points corresponding to each physical quantity;
step 502: and converting the physical quantity corresponding to the binary feature group element and the value of the binary point into an if-then generation formula rule. The purpose of expanding the binary feature group and the binary vector in the binary feature combination optimization model is to cover the direction of the if-then generation formula rule symbol. When n elements before expanding the binarization feature group have '1', the converted production formula rule is that the positive direction is larger than the rule; when n elements have '1' after the binarization feature set is expanded, the converted production formula rule is reverse smaller than the rule.
In one embodiment of the invention, the if-then production rule set for fault F (2) includes 1 set of 7 rules: PV (10) >0.511, PV (11) >80.889, PV (18) <64.022, PV (19) <174.540, PV (22) >78.360, OV (6) >61.121, OV (9) < 34.434; the if-then production rule set for fault F (7) includes 1 set of 1 item rules OV (4) > 69.709; the if-then production rule set for fault F (10) includes 1 set of 1-item rules PV (19) < 216.560. As shown in fig. 2, the normalized distribution of the physical quantities corresponding to the last 80 samples of the fault state and the normal state (the physical quantities are already in the stable state), the normalized rule corresponding to the three faults in fig. 2 is F (2): PV (10) >14.276, PV (11) >3.512, PV (18) < -3.946, PV (19) < -5.509, PV (22) >4.121, OV (6) >14.391, OV (9) < -4.889; f (7): OV (4) > 6.693; f (10): PV (19) < -1.617. It can be seen that the method effectively identifies the key physical quantity for distinguishing the fault state sample from the normal state sample and converts the distribution rule of the key physical quantity into the production rule.
Step 60: performing fault tracing analysis according to the production rule set;
in the step 60, the reason of the fault occurrence is analyzed according to the physical quantity and the corresponding limit value included in the production rule, so that the fault tracing is realized;
in another embodiment of the present specification, the effectiveness of the method of the present invention is demonstrated by taking a lithium ion power battery slurry preparation process as an example. The slurry preparation process (as shown in fig. 3) is the first step of manufacturing the lithium ion power battery, and comprises the steps of mixing the raw materials such as the electrode active powder material, the solvent and the like according to a certain proportion, putting the mixture into a dispersion tank, and driving a stirrer by a stirrer to prepare uniform slurry with good fluidity. Various faults and quality fluctuations may exist in the slurry preparation process, such as equipment faults, sensor faults, excessive slurry viscosity and solid content, caking or air bubbles and the like, and the faults can be directly reflected in process variables, and can be screened and converted into corresponding production rules by the method disclosed by the invention. The process variables of the slurry preparation process include stirring speed, stirring current, vacuum pressure, cooling water valve state, feed liquid temperature and the like. For example, when a slurry high-temperature alarm is generated, data from the beginning of a process to the current time and normal state data in the same time range are directly derived from a database, a production rule is generated by calling the method, the temperature of the slurry can be obviously deviated from the normal state, and meanwhile, the state of the cooling water valve is closed, so that the cooling water valve can be concluded to be in a fault state or not adjusted to be in an automatic opening state; when the viscosity of the slurry deviates from the standard, normal and fault data are derived through the method according to the process flow by stages, so that the condition that the stirring current is larger than the normal state at a certain stirring stage can be seen, the condition that the water injection quantity at the stage cannot meet the requirement can be inferred, the resistance of the stirrer is increased, and the fault reason is judged to be caused by misoperation of workers or the material by further combining with the process standard; when the pressure sensor fails, the method can directly lock the vacuum pressure value abnormity, and further take measures to eliminate the failure.
So far, the technical solutions of the present invention have been described in connection with the preferred embodiments shown in the drawings, but it is easily understood by those skilled in the art that the scope of the present invention is obviously not limited to these specific embodiments. Equivalent changes or substitutions of related technical features can be made by those skilled in the art without departing from the principle of the invention, and the technical scheme after the changes or substitutions can fall into the protection scope of the invention.
Claims (9)
1. A fault tracing method in a lithium ion battery production process is characterized by comprising the following steps:
step 10: acquiring a normal state and fault state sample set in the production process of the lithium ion battery;
step 20: performing data binarization operation on the normal state data and the fault state data in the step 10;
step 30: constructing a binary characteristic combination for distinguishing the normal state data and the fault state data after the binarization operation to optimize a model;
step 40: solving a binary feature group set which enables the model objective function in the binary feature combined optimization model constructed in the step 30 to reach an extreme value;
step 50: reversely converting the binary feature set obtained in the step 40 into a production rule set based on physical quantity;
step 60: and (4) performing fault tracing analysis according to the generated rule set obtained in the step 50.
2. The method for tracing faults as claimed in claim 1, wherein in step 10, the normal state and fault state samples contain physical quantities capable of reflecting the production process of the lithium ion battery, and the physical quantities comprise process variables and control variables; and acquiring physical quantity measured values of a period of continuous time in a normal state and a fault state through a data acquisition system to form a normal state and fault state sample set.
3. The method of fault tracing according to claim 2, wherein step 20 comprises:
step 201: mixing and sorting the normal state physical quantities and the fault state physical quantities, and searching the average value of all adjacent normal state physical quantities and fault state physical quantities as a binarization reference point;
step 202: for each physical quantity, constructing a Gaussian distribution function by using the mean value and the variance of the normal state samples, calculating the number r of the fault state samples in the range of 3 times of the standard deviation of the Gaussian distribution function of the normal state samples, and setting the upper limit and the lower limit T of a number thresholdt、TbTo obtain a binarization point;
step 203: and for each physical quantity, mapping the physical quantity into binary features according to the relative size of the value of the physical quantity and all binary point values, wherein all binary features form a binary vector.
4. The failure tracing method according to claim 3, wherein in step 202, the obtaining of the binary point includes three conditions: when r > TtTaking the maximum value and the minimum value of the binarization reference points of the corresponding physical quantities as binarization points; when T isb≤r≤TtTaking the statistical value of the binary reference point corresponding to the physical quantity as a binary point; when r is less than TbThen, all the binarization reference points corresponding to the physical quantities are taken as the binarization points.
5. The method for tracing faults as claimed in claim 2, wherein in step 30, the binary feature combination optimization model includes three targets and four constraints;
the three targets are: matching degree of the binarization feature group and the normal state sample, difference degree of the binarization feature group and the fault state sample and number of elements in the binarization feature group;
the four constraints include: the number of elements of the binary characteristic group is restricted, the unidirectional restriction of a production formula rule, the matching restriction of normal state samples and the difference restriction of fault state samples,
the model objective function representation is minimized to be the weighted sum of the matching degree of the binarization feature group and the normal state sample, the difference degree of the binarization feature group and the fault state sample and the number of elements in the binarization feature group.
6. The fault tracing method of claim 5, wherein the binary feature combinatorial optimization model is represented as:
where the first row represents the model objective function, min represents the minimization,s、m、d、lall represent the optimization variables of the system,XandYrespectively representing normal state and fault state sample sets, equations (1) - (8) representing constraints of the optimization variables,s.t.representing that the constraint conditions shown in the formulas (1) to (8) need to be satisfied;m i representing a binary feature set withiThe matching degree of the normal state samples is obtained,d i representing a binary feature set withjThe degree of difference of the samples of the individual fault states,lrepresenting the number of elements in the binarized feature set,αandβis the weight of the corresponding target;
is shown to have 2nExtended binary feature set of individual elements, frontnEach element is a binary characteristic group element, and thennIt is the element-wise inverse value of the preceding binarized feature value element,kandprespectively representing a binary feature group number and an element number in the group;
and representing an expanded binary vector, wherein the expansion mode is the same as the binary feature group.
7. The fault tracing method according to claim 2, wherein in step 40, the solving method of the binary feature group set is a branch-and-bound method, a genetic algorithm or a particle swarm algorithm.
8. The fault tracing method of claim 2, wherein step 50 comprises:
step 501: determining the physical quantity corresponding to the binary feature set element according to the number of the binary points corresponding to each physical quantity;
step 502: and converting the physical quantity corresponding to the binary characteristic group element and the value of the binary point into a if-then generating formula rule according to the binary characteristic group element, wherein if a certain physical quantity is greater than or less than a limit value, a certain fault occurs.
9. The method according to claim 8, wherein in step 60, the cause of the fault is analyzed according to the physical quantity included in the production rule and the limit value corresponding to the physical quantity, so as to trace the source of the fault.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011219367.4A CN112099477B (en) | 2020-11-05 | 2020-11-05 | Fault tracing method in lithium ion battery production process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011219367.4A CN112099477B (en) | 2020-11-05 | 2020-11-05 | Fault tracing method in lithium ion battery production process |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112099477A true CN112099477A (en) | 2020-12-18 |
| CN112099477B CN112099477B (en) | 2021-02-02 |
Family
ID=73785472
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011219367.4A Active CN112099477B (en) | 2020-11-05 | 2020-11-05 | Fault tracing method in lithium ion battery production process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112099477B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN202870560U (en) * | 2012-11-20 | 2013-04-10 | 湖南四菱工控技术有限公司 | Lithium battery production line control device with intelligent monitoring system |
| CN106959684A (en) * | 2017-03-29 | 2017-07-18 | 清华大学 | Method for diagnosing faults and device based on compressed sensing and joint norm optimization |
| CN108519479A (en) * | 2018-04-19 | 2018-09-11 | 湖南乐准智芯生物科技有限公司 | A kind of online self-diagnosis system of biochip reaction process |
| CN110262284A (en) * | 2019-06-24 | 2019-09-20 | 江苏科瑞德智控自动化科技有限公司 | A kind of lithium battery pole slice milling train control method and system based on the twin technology of number |
| CN110673577A (en) * | 2019-09-23 | 2020-01-10 | 北京化工大学 | Distributed monitoring and fault diagnosis method for complex chemical production process |
| CN111733498A (en) * | 2020-06-04 | 2020-10-02 | 武汉道非科技有限公司 | Intelligent spinning spun yarn fault detection system |
-
2020
- 2020-11-05 CN CN202011219367.4A patent/CN112099477B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN202870560U (en) * | 2012-11-20 | 2013-04-10 | 湖南四菱工控技术有限公司 | Lithium battery production line control device with intelligent monitoring system |
| CN106959684A (en) * | 2017-03-29 | 2017-07-18 | 清华大学 | Method for diagnosing faults and device based on compressed sensing and joint norm optimization |
| CN108519479A (en) * | 2018-04-19 | 2018-09-11 | 湖南乐准智芯生物科技有限公司 | A kind of online self-diagnosis system of biochip reaction process |
| CN110262284A (en) * | 2019-06-24 | 2019-09-20 | 江苏科瑞德智控自动化科技有限公司 | A kind of lithium battery pole slice milling train control method and system based on the twin technology of number |
| CN110673577A (en) * | 2019-09-23 | 2020-01-10 | 北京化工大学 | Distributed monitoring and fault diagnosis method for complex chemical production process |
| CN111733498A (en) * | 2020-06-04 | 2020-10-02 | 武汉道非科技有限公司 | Intelligent spinning spun yarn fault detection system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112099477B (en) | 2021-02-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chai et al. | Enhanced random forest with concurrent analysis of static and dynamic nodes for industrial fault classification | |
| Ge et al. | Plant-wide industrial process monitoring: A distributed modeling framework | |
| CN110213222B (en) | Network intrusion detection method based on machine learning | |
| CN111782472A (en) | System abnormality detection method, device, equipment and storage medium | |
| CN105677791B (en) | For analyzing the method and system of the operation data of wind power generating set | |
| CN110297207A (en) | Method for diagnosing faults, system and the electronic device of intelligent electric meter | |
| CN111639304B (en) | CSTR fault positioning method based on Xgboost regression model | |
| CN113408659A (en) | Building energy consumption integrated analysis method based on data mining | |
| CN114114039A (en) | Method and device for evaluating consistency of single battery cells of battery system | |
| CN112099477B (en) | Fault tracing method in lithium ion battery production process | |
| Ma et al. | A hybrid fault diagnosis method based on fuzzy signed directed graph and neighborhood rough set | |
| CN115146727A (en) | Intelligent power utilization system fault detection method and system | |
| CN113689036A (en) | Thermal imager quality problem reason prediction method based on decision tree C4.5 algorithm | |
| CN113726558A (en) | Network equipment flow prediction system based on random forest algorithm | |
| CN111881040A (en) | Test data generation method of abstract state model based on recurrent neural network | |
| CN112016597A (en) | Depth sampling method based on Bayesian imbalance measurement in machine learning | |
| Verma et al. | Deep learning-based fault detection in the Tennessee Eastman process | |
| Wang et al. | Multimode process fault detection method based on variable local outlier factor | |
| CN110736888A (en) | method for monitoring abnormal electricity consumption behavior of user | |
| Peng et al. | A condition monitoring approach based on isolation forest algorithm for nuclear power plant | |
| Weng et al. | A Correlation Analysis-Based Multivariate Alarm Method With Maximum Likelihood Evidential Reasoning | |
| Zhang et al. | Performance assessment of fault classifier of chemical plant based on support vector machine | |
| CN117633722B (en) | Detection control method and system based on intelligent detection robot | |
| CN113158552B (en) | Bioreactor operation condition grading prediction method and system based on time sequence | |
| Yélamos et al. | Fault diagnosis based on support vector machines and systematic comparison to existing approaches |
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 |