CN110990941B - DPF parameter setting method, system and storage medium - Google Patents
DPF parameter setting method, system and storage medium Download PDFInfo
- Publication number
- CN110990941B CN110990941B CN201911047575.8A CN201911047575A CN110990941B CN 110990941 B CN110990941 B CN 110990941B CN 201911047575 A CN201911047575 A CN 201911047575A CN 110990941 B CN110990941 B CN 110990941B
- Authority
- CN
- China
- Prior art keywords
- position data
- particle position
- dpf
- particle
- last
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Processes For Solid Components From Exhaust (AREA)
Abstract
The embodiment of the invention discloses a DPF parameter setting method, a DPF parameter setting system and a storage medium, wherein the DPF parameter setting method comprises the following steps: substituting the particle position data obtained in each time into a pre-constructed fitness function to obtain a fitness value corresponding to each particle position data; dividing the particle position data into a last particle position data set and a non-last particle position data set according to the fitness value; acquiring the j-th global optimal particle position data and an adaptability value corresponding to the j-th global optimal particle position data; when the preset iteration times are not reached, carrying out iterative updating on the particle positions in the last particle position data set according to a second preset rule, carrying out iterative updating on the particle positions in the non-last particle position data according to a third preset rule, and repeating the steps as the next pre-acquired particle position data; until reaching the preset iteration times; and selecting the optimal particle position data from all the global optimal particle position data, and taking the optimal particle position data as a numerical value corresponding to the parameter to be set of the particle catcher DPF.
Description
Technical Field
The embodiment of the invention relates to the technical field of data processing, in particular to a method, a system and a storage medium for setting parameters of a particle catcher (Diesel Particulate Filter, DPF for short).
Background
Along with the rapid upgrading of the economy and the gradual improvement of the requirements on living environment in China, the reduction of the pollution degree of motor vehicles to the atmosphere environment is a necessary trend of development in China. With the proposal of the national six-emission standard, the DPF becomes a necessary device for diesel vehicles, especially for load-carrying vehicles, and has good development prospect.
DPF is a ceramic filter installed in the exhaust system of diesel engine, which can catch the particulate matters in the exhaust gas of diesel engine before it is discharged to the atmosphere, and related researches show that DPF can reduce the soot generated by diesel engine by more than about 90%, and can effectively reduce the pollution degree of diesel engine to the atmosphere environment when working.
However, for development of the DPF, if the physical equipment development is adopted, a large amount of development cost is required. The simulation technology can accelerate the research and development speed and reduce the research and development cost. However, some parameters which are difficult to set often exist in the simulation model, and the credibility of the simulation model is affected.
Therefore, how to set the parameters difficult to set in the simulation model, so as to reduce the average value of the absolute value of the error between the simulation model output and the corresponding bench test data to the greatest extent becomes the technical problem to be solved urgently.
Disclosure of Invention
Therefore, the embodiment of the invention provides a DPF parameter setting method, a DPF parameter setting system and a DPF parameter storage medium, which are used for solving the technical problem that the average value of the absolute value of the error between the output of a simulation model and corresponding bench test data is minimized to be urgently solved in the application.
In order to achieve the above object, the embodiment of the present invention provides the following technical solutions:
according to a first aspect of an embodiment of the present invention, there is provided a DPF parameter tuning method, including:
substituting each particle position data in the j-th pre-acquired particle position data into a pre-constructed fitness function respectively to acquire a fitness value corresponding to each particle position data;
dividing all the particle position data into a last particle position data set and a non-last particle position data set according to a first preset rule according to the fitness value corresponding to each particle position data;
acquiring the j-th global optimal particle position data and an adaptability value corresponding to the j-th global optimal particle from a non-last particle data set;
when the value of j does not reach the value corresponding to the preset iteration times, carrying out iterative updating on the particle positions in the last particle position data set according to a second preset rule, and carrying out iterative updating on the particle positions in the non-last particle position data according to a third preset rule;
Taking all the particle position data subjected to iterative updating as j+1th pre-acquired particle position data, respectively inputting the j+1th pre-acquired particle position data into a pre-constructed fitness function, and acquiring fitness values corresponding to the j+1th pre-acquired particle position data;
dividing all the j+1th pre-acquired particle position data into a last particle position data set and a non-last particle position data set according to a first preset rule according to the fitness value corresponding to each j+1th pre-acquired particle position data;
acquiring j+1th global optimal particle position data and a fitness value corresponding to the j+1th global optimal particle from the j+1th determined non-last particle position data set;
acquiring last global optimal particle position data and an adaptability value corresponding to the last global optimal particle from the last acquired non-last particle position data set until the iteration number reaches the preset iteration number;
selecting optimal particle position data from all the acquired global optimal particle position data according to the fitness value corresponding to each global optimal particle, and taking the optimal particle position data as a numerical value corresponding to a DPF to-be-set parameter of the particle catcher for the last time; j is a numerical value which is larger than or equal to 1 and smaller than or equal to a numerical value corresponding to preset iteration times, the initial value of j is 1, the values are sequentially transmitted, the 1 st particle position data are initialized particle position data corresponding to DPF parameters to be set, and the dimension of each particle position data is the same as the number of the DPF parameters to be set.
Further, according to the fitness value corresponding to each particle position data, dividing all the particle position data into a last particle position data set and a non-last particle position data set according to a first preset rule, specifically including:
sequencing and numbering all the particle position data according to the corresponding fitness value of each particle position data;
all particle position data are divided into a last particle position data set and a non-last particle position data set by number.
Further, performing iterative updating on the particle positions in the last particle position data set according to a second preset rule, and performing iterative updating on the particle positions in the non-last particle position data according to a third preset rule, specifically including:
the particle positions in the last particle position dataset are iteratively updated according to equation 1,
x i,j+1 ＝x m1 -rand·(x m1 -x m2 ),N tail ≤i≤N，1≤j<t (equation 1)
The particle positions in the non-last particle position data are iteratively updated according to equation 2,
x i,j+1 ＝wx i,j +c1·rand·(x m3 -x i,j ),1≤i<N tail ，1≤j<t (formula 2)
Wherein i is the position number of the particles after sorting; j is the current iteration number; j+1 is the next iteration number; t is the preset iteration times; w is the inertia coefficient; ntail is the serial number of the last particle position with the best adaptability; m1 and m2 are the position numbers of any two non-terminal particles, namely, 1.ltoreq.m2 < m1< Ntail; m3 is any particle position number with a number smaller than i; c1 is a learning factor; rand is any random value within interval 0, 1.
Further, before substituting each particle position data in the particle position data pre-acquired in any iteration to the pre-constructed fitness function, the method further includes:
when the first particle position data is out of the preset value range, the first particle position data is corrected according to the formula 3,
x i,j+1 ＝X min +rand·(X max -X min ) (equation 3)
Wherein i is the ordered particlesA position number; j is the current iteration number; j+1 is the next iteration number; x is X min Presetting a lower limit of a range for the particle position; x is X max Presetting an upper limit of a range for the particle position; rand is arbitrarily at [0,1]Random numbers in the interval, and the first particle position data is any particle position data.
Further, the parameters to be set for the DPF include:
the exhaust gas volume flow PT1 filter coefficient, the flow resistance PT1 filter coefficient and the DPF inlet exhaust gas volume PT1 filter coefficient in the regeneration mode.
Further, according to the fitness value corresponding to each global optimal particle, selecting optimal particle position data from all acquired global optimal particle position data, and the method further comprises:
inputting the optimal particle position data and the pre-acquired DPF experimental input data into a pre-constructed DPF simulation model to acquire DPF output simulation data;
And calculating an absolute difference value between the output simulation data and the DPF experimental output data by using the objective function.
Further, the DPF experimental input data includes: the exhaust gas volume flow calculation value, the DPF surface temperature, the engine running state signal, the engine running time, the engine current running mode signal, the DPF pressure difference, the DPF function activation signal, the ash correction coefficient, the DPF carbon loading, the discontinuous DPF carbon loading simulation value, the DPF simulation state activation signal, the ash volume, the regeneration requirement signal, the regeneration incomplete signal, the DPF simulation state activation state word, the environment temperature, the environment pressure and the pressure signal are not trusted signals;
the DPF experimental output data includes: carbon loading measured value, flow resistance after filtration in DPF, exhaust gas mass flow, carbon loading base value.
Further, the objective function is expressed by equation 4:
wherein sim is a,b B-th discrete data contained in a-th simulation result data in the optimal particle position data; real a,b Represents the b-th discrete data contained in the a-th experimental output data in the DPF experimental output data, wherein sim a,b And real a,b There is a mapping relationship between them.
According to a second aspect of an embodiment of the present invention, there is provided a DPF parameter tuning system including: a processor and a memory;
The memory is used for storing one or more program instructions;
a processor for executing one or more program instructions for performing any of the method steps of the DPF parameter tuning method as described above.
According to a third aspect of embodiments of the present invention, there is provided a computer storage medium having one or more program instructions embodied therein for performing any of the above-described DPF parameter tuning methods by a DPF parameter tuning system.
The embodiment of the invention has the following advantages: and substituting the pre-acquired particle position data after each iteration into a pre-constructed fitness function respectively, and acquiring a fitness value corresponding to each particle position data, wherein the iterative updating is not directly performed. The particle position data are sorted according to the fitness value corresponding to the particle position data, and then two groups are divided. One of the sets is a last particle position data set and one of the sets is a non-last particle position data set. And acquiring the j-th global optimal particle position data and the fitness value corresponding to the j-th global optimal particle from the non-last particle data set. And updating the two groups of data according to different rules respectively to serve as the data input in the next iteration. And acquiring and recording the j+1th global optimal particle position data, the fitness value corresponding to the j+1th global optimal particle and the fitness value corresponding to the j+1th global optimal particle from the j+1th determined non-last particle position data set. Ending the iteration time when the iteration time reaches the preset iteration time, and obtaining the last global optimal particle position data and the fitness value corresponding to the last global optimal particle in the non-last particle position data set. And selecting optimal particle position data from all the acquired global optimal particle position data according to the fitness value corresponding to each global optimal particle, and taking the optimal particle position data as a numerical value corresponding to the parameter to be set of the particle catcher DPF. In this way, the parameters to be set of the DPF are set by the improved particle swarm algorithm with the final elimination mechanism, so that the average value of the absolute value of the error between the simulation model output and the corresponding bench test data is reduced to the greatest extent.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below. It will be apparent to those skilled in the art from this disclosure that the drawings described below are merely exemplary and that other embodiments may be derived from the drawings provided without undue effort.
The structures, proportions, sizes, etc. shown in the present specification are shown only for the purposes of illustration and description, and are not intended to limit the scope of the invention, which is defined by the claims, so that any structural modifications, changes in proportions, or adjustments of sizes, which do not affect the efficacy or the achievement of the present invention, should fall within the ambit of the technical disclosure.
FIG. 1 is a schematic flow chart of a DPF parameter setting method provided in embodiment 1 of the present invention;
FIG. 2 is a schematic structural diagram of a DPF parameter setting device according to embodiment 2 of the present invention;
fig. 3 is a schematic structural diagram of a DPF parameter tuning system according to embodiment 3 of the present invention.
Detailed Description
Other advantages and advantages of the present invention will become apparent to those skilled in the art from the following detailed description, which, by way of illustration, is to be read in connection with certain specific embodiments, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Embodiment 1 of the present invention provides a method for setting parameters of a DPF, specifically, as shown in fig. 1, before executing the method of the embodiment of the present invention, some preparation work is first performed. Such as initializing some data. For example, initializing certain parameters in the particle swarm algorithm, and manually setting the total number of the individual particles, the maximum iteration number, the weight coefficient, the learning factor, the particle position value range and the like.
The particle position data is initialized, for example, a random function is used to randomly generate N primary particle positions with D dimensions in the range of values of the dimension variables. The value of D here is equal to the number of parameters to be set for the DPF. In practice, each particle position data is formed by an array of dimension D, optionally with elements in the array ranging from 0, 100. Optionally, the parameters to be set for the DPF include: the exhaust gas volume flow PT1 filter coefficient, the flow resistance PT1 filter coefficient and the DPF inlet exhaust gas volume PT1 filter coefficient in the regeneration mode.
The method comprises the following steps:
step 110, substituting each particle position data in the j-th pre-acquired particle position data into the pre-constructed fitness function, and acquiring a fitness value corresponding to each particle position data.
Specifically, j is a value greater than or equal to 1 and less than or equal to a value corresponding to a preset iteration number, and the initial value of j is 1 and sequentially progressive values. When j is equal to 1, i.e. the first pre-acquired particle position data is the initialized particle position data described hereinabove.
The pre-constructed fitness function is the fitness function in the particle swarm algorithm described above. And each iteration is performed once, the iteration times in the particle swarm algorithm are automatically increased by 1. After each fitness value corresponding to each particle position data pre-acquired for the j-th time is acquired, step 120 is performed.
And 120, dividing all the particle position data into a last particle position data set and a non-last particle position data set according to a first preset rule according to the fitness value corresponding to each particle position data.
Specifically, all the particle position data can be ordered and numbered according to the corresponding fitness value of each particle position data; all particle position data are then divided into a last particle position data set and a non-last particle position data set by number.
For example, the particle position data includes 10000, and particle position data numbered 1 to 5000 belong to a non-last particle data group, and particle position data numbered 5001 to 10000 belong to a last particle data group.
And 130, acquiring and recording the j-th global optimal particle position data and the fitness value corresponding to the j-th global optimal particle from the non-last particle data set.
The specific algorithm can be referred to the principle of the particle swarm algorithm, and will not be described in detail herein.
And 140, when the value of j does not reach the value corresponding to the preset iteration times, carrying out iterative updating on the particle positions in the last particle position data set according to a second preset rule, and carrying out iterative updating on the particle positions in the non-last particle position data according to a third preset rule.
In the above description, once for each iteration, the number of iterations in the particle swarm algorithm is automatically increased by 1, and when the value of j does not reach the value corresponding to the preset number of iterations (for example, 1000 times), then the particle positions in the last particle position data set are iteratively updated according to the second preset rule, and the particle positions in the non-last particle position data set are iteratively updated according to the third preset rule.
Alternatively, the particle positions in the last particle position data set are iteratively updated, for example using the following formula,
x i,j+1 ＝x m1 -rand·(x m1 -x m2 ),N tail ≤i≤N，1≤j<t (equation 1)
The particle positions in the non-last particle position data are iteratively updated using the following formula,
x i,j+1 ＝wx i,j +c1·rand·(x m3 -x i,j ),1≤i<N tail ，1≤j<t (formula 2)
Wherein i is the position number of the particles after sorting; j is the current iteration number; j+1 is the next iteration number; t is the preset iteration times; w is the inertia coefficient; ntail is the serial number of the last particle position with the best adaptability; m1 and m2 are the position numbers of any two non-terminal particles, namely, 1.ltoreq.m2 < m1< Ntail; m3 is any particle position number with a number smaller than i; c1 is a learning factor; rand is any random value within interval 0, 1.
After the particle position data is iteratively updated in the manner of step 140, the next pre-acquired particle position data is used, and then the steps 110 to 130 are repeatedly performed.
That is, steps 150 to 170 are performed.
And 150, substituting all the particle position data subjected to iterative updating as the j+1th pre-acquired particle position data into a pre-constructed fitness function respectively, and acquiring fitness values corresponding to the j+1th pre-acquired particle position data.
Step 160, dividing all the j+1th pre-acquired particle position data into a last particle position data set and a non-last particle position data set according to a first preset rule according to the fitness value corresponding to each j+1th pre-acquired particle position data.
Step 170, acquiring and recording j+1th global optimal particle position data and fitness value corresponding to j+1th global optimal particles from j+1th determined non-last particle position data set.
And 180, acquiring the last global optimal particle position data and the fitness value corresponding to the last global optimal particle from the last acquired non-last particle position data set until the iteration number reaches the preset iteration number.
The specific procedure of steps 150 to 180 is similar to the procedure already described above and will not be described here too much.
And 190, selecting optimal particle position data from all the acquired global optimal particle position data according to the fitness value corresponding to each global optimal particle, and taking the optimal particle position data as a numerical value corresponding to the parameter to be set of the particle catcher DPF.
Specifically, the best fitness is selected from the global optimal particle position data, namely the optimal particle position data.
Wherein, all data elements in the optimal particle position data are actually the values of all corresponding parameters to be set. The global optimal particle position data acquired in each iteration process is actually the particle position data with the first order (numbered 1) in the non-last particle position data set, because the better the fitness is, the earlier the ordering is. And selecting optimal particle position data from all global optimal particle position data, wherein an example corresponding to the optimal particle position data is the optimal particle, the position data is denoted as Prest, and the fitness is denoted as Fbest.
Further optionally, to ensure accuracy of the data input into the fitness function, the method may further include: and before substituting each particle position data in the particle position data pre-acquired in any iteration into a pre-constructed fitness function, correcting the data of which the particle position data value is out of a preset value range. The preset value range here may preferably be [0,100] as described above.
Therefore, each particle position data needs to be examined, and once it is determined that a certain particle position data is out of a preset value range, the following formula can be adopted for correction:
x i,j+1 ＝X min +rand·(X max -X min ) (equation 3)
Wherein i isPosition number of the particles after sequencing; j is the current iteration number; j+1 is the next iteration number; x is X min Presetting a lower limit of a range for the particle position; x is X max Presetting an upper limit of a range for the particle position; rand is arbitrarily at [0,1]Random numbers in the interval, and the first particle position data is any particle position data.
Optionally, the parameters to be set have been described above, and will not be described here again. And after the last global optimum particle position data is obtained from the last obtained non-last particle position data set, the method may further include:
inputting the optimal particle position data and the pre-acquired DPF experimental input data into a pre-constructed DPF simulation model to acquire DPF output simulation data;
and calculating an absolute difference value between the output simulation data and the DPF experimental output data by using the objective function.
Taking diesel engine DPF of some automobile electronics company as an example, acquisition time length is 1600s, sampling period is 10ms, and 160000 groups of DPF experiment input data and DPF experiment output data are obtained. DPF experiment input data and DPF experiment output data are obtained through bench calibration tests.
The DPF experimental input data includes: the exhaust gas volume flow calculation value, the DPF surface temperature, the engine running state signal, the engine running time, the engine current running mode signal, the DPF pressure difference, the DPF function activation signal, the ash correction coefficient, the DPF carbon loading, the discontinuous DPF carbon loading simulation value, the DPF simulation state activation signal, the ash volume, the regeneration requirement signal, the regeneration incomplete signal, the DPF simulation state activation state word, the environment temperature, the environment pressure and the pressure signal are not trusted signals;
The DPF experimental output data includes: carbon loading measured value, flow resistance after filtration in DPF, exhaust gas mass flow, carbon loading base value.
And establishing a DPF simulation model based on the 160000 groups of parameter data acquired and preparing for optimizing DPF simulation parameters.
In this example, the DPF simulation model may be built in a MATLAB/Simulink environment from MathWorks, inc. of America. Of course, the simulation may also be implemented using other simulation software known. And obtaining DPF output simulation data through a simulation model.
Then, an absolute difference between the output simulation data and the DPF experimental output data is calculated using the objective function. Wherein the objective function is expressed by equation 4:
wherein sim is a,b B-th discrete data contained in a-th simulation result data in the optimal particle position data; real a,b Represents the b-th discrete data contained in the a-th experimental output data in the DPF experimental output data, wherein sim a,b And real a,b There is a mapping relationship between them.
Through the simulation and calculation of the process, the obtained optimal exhaust gas volume flow PT1 filter coefficient, the flow resistance PT1 filter coefficient and the flow resistance PT1 filter coefficient in the regeneration mode and the DPF inlet exhaust gas volume PT1 filter coefficient are 6.4191, 1.0925, 33.5999, 0.3759 and 2.7769. Compared with the original particle swarm algorithm, the average value of the carbon load measured value, the flow resistance after filtration in the DPF, the exhaust gas mass flow and the relative error absolute value of the carbon load basic value output by the DPF simulation model optimized by the improved particle swarm algorithm with the final-position elimination mechanism introduced by the implementation is respectively reduced by 36.38%, 39.68%, 51.76% and 37.22%.
According to the DPF parameter setting method provided by the embodiment of the invention, after particle position data pre-acquired after each iteration are respectively substituted into a pre-constructed fitness function, the fitness value corresponding to each particle position data is acquired, and then the iteration update is not directly performed. The particle position data are sorted according to the fitness value corresponding to the particle position data, and then two groups are divided. One of the sets is a last particle position data set and one of the sets is a non-last particle position data set. And acquiring the j-th global optimal particle position data and the fitness value corresponding to the j-th global optimal particle from the non-last particle data set. And updating the two groups of data according to different rules respectively to serve as the data input in the next iteration. And acquiring and recording the j+1th global optimal particle position data, the fitness value corresponding to the j+1th global optimal particle and the fitness value corresponding to the j+1th global optimal particle from the j+1th determined non-last particle position data set. Ending the iteration time when the iteration time reaches the preset iteration time, and obtaining the last global optimal particle position data and the fitness value corresponding to the last global optimal particle in the non-last particle position data set. And selecting optimal particle position data from all the acquired global optimal particle position data according to the fitness value corresponding to each global optimal particle, and taking the optimal particle position data as a numerical value corresponding to the parameter to be set of the particle catcher DPF. In this way, the parameters to be set of the DPF are set by the improved particle swarm algorithm with the final elimination mechanism, so that the average value of the absolute value of the error between the simulation model output and the corresponding bench test data is reduced to the greatest extent.
Corresponding to the above embodiment 1, embodiment 2 of the present invention further provides a DPF parameter setting device, specifically as shown in fig. 2, which includes: processing unit 201, grouping unit 202, and update iteration unit 203.
A processing unit 201, configured to substitute each particle position data in the jth pre-acquired particle position data into a pre-constructed fitness function, to acquire a fitness value corresponding to each particle position data;
a grouping unit 202, configured to divide all the particle position data into a last particle position data set and a non-last particle position data set according to a first preset rule according to the fitness value corresponding to each particle position data;
the processing unit 201 is further configured to acquire and record, from the non-last particle data set, jth global optimal particle position data and an fitness value corresponding to the jth global optimal particle;
an updating iteration unit 203, configured to iteratively update the particle positions in the last particle position data set according to a second preset rule and iteratively update the particle positions in the non-last particle position data according to a third preset rule when the value of j does not reach the value corresponding to the preset iteration number;
The processing unit 201 is further configured to replace all the iteratively updated particle position data as j+1th pre-acquired particle position data into a pre-constructed fitness function, to obtain fitness values corresponding to each j+1th pre-acquired particle position data;
the grouping unit 202 is further configured to divide all the j+1th pre-acquired particle position data into a last particle position data set and a non-last particle position data set according to a first preset rule according to the fitness value corresponding to each j+1th pre-acquired particle position data;
the processing unit 201 is further configured to acquire and record the j+1th global optimal particle position data and an fitness value corresponding to the j+1th global optimal particle from the j+1th determined non-last particle position data set;
acquiring last global optimal particle position data and an adaptability value corresponding to the last global optimal particle from the last acquired non-last particle position data set until the iteration number reaches the preset iteration number;
selecting optimal particle position data from all the acquired global optimal particle position data according to the fitness value corresponding to each global optimal particle, and taking the optimal particle position data as a numerical value corresponding to a DPF to-be-set parameter of the particle catcher for the last time; j is a numerical value which is larger than or equal to 1 and smaller than or equal to a numerical value corresponding to preset iteration times, the initial value of j is 1, the values are sequentially transmitted, the 1 st particle position data are initialized particle position data corresponding to DPF parameters to be set, and the dimension of each particle position data is the same as the number of the DPF parameters to be set.
Optionally, the grouping unit 202 is specifically configured to sort and number all the particle position data according to the fitness value corresponding to each particle position data;
all particle position data are divided into a last particle position data set and a non-last particle position data set by number.
Optionally, the updating iteration unit 203 is specifically configured to:
the particle positions in the last particle position dataset are iteratively updated according to equation 1,
x i,j+1 ＝x m1 -rand·(x m1 -x m2 ),N tail ≤i≤N，1≤j<t (equation 1)
The particle positions in the non-last particle position data are iteratively updated according to equation 2,
x i,j+1 ＝wx i +c1·rand·(x m3 -x i ),1≤i<N tail ，1≤j<t (formula 2)
Wherein i is the position number of the particles after sorting; j is the current iteration number; j+1 is the next iteration number; t is the preset iteration times; w is the inertia coefficient; ntail is the serial number of the last particle position with the best adaptability; m1 and m2 are the position numbers of any two non-terminal particles, namely, 1.ltoreq.m2 < m1< Ntail; m3 is any particle position number with a number smaller than i; c1 is a learning factor; rand is any random value within interval 0, 1.
Optionally, the apparatus further comprises: a correction unit 204 for correcting the first particle position data according to formula 3 when the first particle position data is outside the preset value range,
x i,j+1 ＝X min +rand·(X max -X min ) (equation 3)
Wherein i is the position number of the particles after sorting; j is the current iteration number; j+1 is the next iteration number; x is X min Presetting a lower limit of a range for the particle position; x is X max Presetting an upper limit of a range for the particle position; rand is arbitrarily at [0,1]Random numbers in the interval, and the first particle position data is any particle position data.
Optionally, the parameters to be set for the DPF include:
the exhaust gas volume flow PT1 filter coefficient, the flow resistance PT1 filter coefficient and the DPF inlet exhaust gas volume PT1 filter coefficient in the regeneration mode.
Optionally, the apparatus further comprises: the simulation unit 205 is configured to input the optimal particle position data and the pre-acquired DPF experimental input data into a pre-constructed DPF simulation model, and acquire DPF output simulation data;
and calculating an absolute difference value between the output simulation data and the DPF experimental output data by using the objective function.
Optionally, the DPF experimental input data includes: the exhaust gas volume flow calculation value, the DPF surface temperature, the engine running state signal, the engine running time, the engine current running mode signal, the DPF pressure difference, the DPF function activation signal, the ash correction coefficient, the DPF carbon loading, the discontinuous DPF carbon loading simulation value, the DPF simulation state activation signal, the ash volume, the regeneration requirement signal, the regeneration incomplete signal, the DPF simulation state activation state word, the environment temperature, the environment pressure and the pressure signal are not trusted signals;
The DPF experimental output data includes: carbon loading measured value, flow resistance after filtration in DPF, exhaust gas mass flow, carbon loading base value.
Alternatively, the objective function is expressed using equation 4:
wherein sim is a,b B-th discrete data contained in a-th simulation result data in the optimal particle position data; real a,b Represents the b-th discrete data contained in the a-th experimental output data in the DPF experimental output data, wherein sim a,b And real a,b There is a mapping relationship between them.
The functions performed by each component in the DPF parameter tuning device provided in the embodiment of the present invention are described in detail in the above embodiment 1, so that redundant description is omitted here.
According to the DPF parameter setting device provided by the embodiment of the invention, after particle position data pre-acquired after each iteration are respectively substituted into a pre-constructed fitness function, the fitness value corresponding to each particle position data is acquired, and then the iteration update is not directly performed. The particle position data are sorted according to the fitness value corresponding to the particle position data, and then two groups are divided. One of the sets is a last particle position data set and one of the sets is a non-last particle position data set. And acquiring the j-th global optimal particle position data and the fitness value corresponding to the j-th global optimal particle from the non-last particle data set. And updating the two groups of data according to different rules respectively to serve as the data input in the next iteration. And acquiring and recording the j+1th global optimal particle position data, the fitness value corresponding to the j+1th global optimal particle and the fitness value corresponding to the j+1th global optimal particle from the j+1th determined non-last particle position data set. Ending the iteration time when the iteration time reaches the preset iteration time, and obtaining the last global optimal particle position data and the fitness value corresponding to the last global optimal particle in the non-last particle position data set. And selecting optimal particle position data from all the acquired global optimal particle position data according to the fitness value corresponding to each global optimal particle, and taking the optimal particle position data as a numerical value corresponding to the parameter to be set of the particle catcher DPF. In this way, the parameters to be set of the DPF are set by the improved particle swarm algorithm with the final elimination mechanism, so that the average value of the absolute value of the error between the simulation model output and the corresponding bench test data is reduced to the greatest extent.
Corresponding to the above embodiment, embodiment 3 of the present invention further provides a DPF parameter tuning system, specifically as shown in fig. 3, which includes: a processor 301 and a memory 302;
memory 302 is used to store one or more program instructions;
the processor 301 is configured to execute one or more program instructions for performing any of the method steps of a DPF parameter tuning method as described in the above embodiments.
According to the DPF parameter setting system provided by the embodiment of the invention, after particle position data pre-acquired after each iteration are respectively substituted into a pre-constructed fitness function, the fitness value corresponding to each particle position data is acquired, and then the iteration update is not directly performed. The particle position data are sorted according to the fitness value corresponding to the particle position data, and then two groups are divided. One of the sets is a last particle position data set and one of the sets is a non-last particle position data set. And acquiring the j-th global optimal particle position data and the fitness value corresponding to the j-th global optimal particle from the non-last particle data set. And updating the two groups of data according to different rules respectively to serve as the data input in the next iteration. And acquiring and recording the j+1th global optimal particle position data, the fitness value corresponding to the j+1th global optimal particle and the fitness value corresponding to the j+1th global optimal particle from the j+1th determined non-last particle position data set. Ending the iteration time when the iteration time reaches the preset iteration time, and obtaining the last global optimal particle position data and the fitness value corresponding to the last global optimal particle in the non-last particle position data set. And selecting optimal particle position data from all the acquired global optimal particle position data according to the fitness value corresponding to each global optimal particle, and taking the optimal particle position data as a numerical value corresponding to the parameter to be set of the particle catcher DPF. In this way, the parameters to be set of the DPF are set by the improved particle swarm algorithm with the final elimination mechanism, so that the average value of the absolute value of the error between the simulation model output and the corresponding bench test data is reduced to the greatest extent.
Corresponding to the above embodiments, the present invention further provides a computer storage medium, which contains one or more program instructions. Wherein the one or more program instructions are for performing a DPF parameter tuning method as described above by a DPF parameter tuning system.
In the embodiment of the invention, the processor may be an integrated circuit chip with signal processing capability. The processor may be a general purpose processor, a digital signal processor (Digital Signal Processor, DSP for short), an application specific integrated circuit (Application Specific f ntegrated Circuit ASIC for short), a field programmable gate array (FieldProgrammable Gate Array FPGA for short), or other programmable logic device, discrete gate or transistor logic device, discrete hardware components.
The disclosed methods, steps, and logic blocks in the embodiments of the present invention may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of the method disclosed in connection with the embodiments of the present invention may be embodied directly in the execution of a hardware decoding processor, or in the execution of a combination of hardware and software modules in a decoding processor. The software modules may be located in a random access memory, flash memory, read only memory, programmable read only memory, or electrically erasable programmable memory, registers, etc. as well known in the art. The processor reads the information in the storage medium and, in combination with its hardware, performs the steps of the above method.
The storage medium may be memory, for example, may be volatile memory or nonvolatile memory, or may include both volatile and nonvolatile memory.
The nonvolatile Memory may be a Read-Only Memory (ROM), a Programmable ROM (PROM), an Erasable PROM (EPROM), an electrically Erasable ROM (Electrically EPROM, EEPROM), or a flash Memory.
The volatile memory may be a random access memory (Random Access Memory, RAM for short) which acts as an external cache. By way of example, and not limitation, many forms of RAM are available, such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (Double Data RateSDRAM), enhanced SDRAM (ESDRAM), synchronous DRAM (SLDRAM), and direct memory bus RAM (directracram, DRRAM).
The storage media described in embodiments of the present invention are intended to comprise, without being limited to, these and any other suitable types of memory.
Those skilled in the art will appreciate that in one or more of the examples described above, the functions described in the present invention may be implemented in a combination of hardware and software. When the software is applied, the corresponding functions may be stored in a computer-readable medium or transmitted as one or more instructions or code on the computer-readable medium. Computer-readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A storage media may be any available media that can be accessed by a general purpose or special purpose computer.
The foregoing detailed description of the invention has been presented for purposes of illustration and description, and it should be understood that the foregoing is by way of illustration and description only, and is not intended to limit the scope of the invention.
Claims (9)
1. A method of DPF parameter tuning, the method comprising:
substituting each particle position data in the j-th pre-acquired particle position data into a pre-constructed fitness function respectively to acquire a fitness value corresponding to each particle position data;
dividing all the particle position data into a last particle position data set and a non-last particle position data set according to a first preset rule according to the fitness value corresponding to each particle position data;
acquiring and recording the j-th global optimal particle position data and the fitness value corresponding to the j-th global optimal particle from the non-last particle data set;
when the value of j does not reach the value corresponding to the preset iteration times, carrying out iterative updating on the particle positions in the last particle position data set according to a second preset rule, and carrying out iterative updating on the particle positions in the non-last particle position data according to a third preset rule, wherein the method specifically comprises the following steps:
the particle positions in the last particle position data set are iteratively updated according to equation 1,
x i，j+1 ＝x m1 -rand·(x m1 -x m2 )，N tail i is not less than N,1 is not less than j is not less than T (formula 1)
Iteratively updating the particle positions in the non-last particle position data according to equation 2,
x i，j+1 ＝wx i，j +c1·rand·(x m3 -x i，j )，1≤i＜N tail J < T (formula 2) of 1 ≡j
Wherein i is the position number of the particles after sorting; j is the current iteration number; j+1 is the next iteration number; t is the preset iteration times; w is the inertia coefficient; n (N) tail The serial number of the last particle position with the best adaptability; m1 and m2 are the position numbers of any two non-terminal particles, i.e. 1.ltoreq.m2 < m1 < N tail The method comprises the steps of carrying out a first treatment on the surface of the m3 is any particle position number with a number smaller than i; c1 is a learning factor; rand is interval [0,1 ]]Any random value within;
taking all the particle position data subjected to iterative updating as j+1th pre-acquired particle position data, respectively inputting the j+1th pre-acquired particle position data into the pre-constructed fitness function, and acquiring fitness values corresponding to the j+1th pre-acquired particle position data;
dividing all the j+1th pre-acquired particle position data into a last particle position data set and a non-last particle position data set according to a first preset rule according to the fitness value corresponding to each j+1th pre-acquired particle position data;
acquiring and recording j+1th global optimal particle position data and an adaptability value corresponding to the j+1th global optimal particle from the j+1th determined non-last particle position data set;
Acquiring last global optimal particle position data and an adaptability value corresponding to the last global optimal particle from a last acquired non-last particle position data set until the iteration number reaches the preset iteration number;
selecting optimal particle position data from all the acquired global optimal particle position data according to the fitness value corresponding to each global optimal particle, and taking the optimal particle position data as a numerical value corresponding to a DPF to-be-set parameter of the particle catcher for the last time; j is a numerical value which is larger than or equal to 1 and smaller than or equal to a numerical value corresponding to preset iteration times, the initial value of j is 1, the values are sequentially transmitted, the 1 st particle position data are initialized particle position data corresponding to DPF parameters to be set, and the dimension of each particle position data is the same as the number of the DPF parameters to be set.
2. The method according to claim 1, wherein the dividing all the particle position data into a last particle position data set and a non-last particle position data set according to the fitness value corresponding to each particle position data according to the first preset rule specifically includes:
sequencing and numbering all the particle position data according to the corresponding fitness value of each particle position data;
All particle position data are divided into a last particle position data set and a non-last particle position data set according to the numbering.
3. The method according to claim 1 or 2, wherein each of the particle position data pre-acquired at any one iteration is substituted into the pre-constructed fitness function before each particle position data is substituted into the fitness function, the method further comprising:
when the first particle position data is out of the preset value range, correcting the first particle position data according to a formula 3,
x i，j+1 ＝X min +rand·(X max -X min ) (equation 3)
Wherein i is the orderPosition number of the rear particle; j is the current iteration number; j+1 is the next iteration number; x is X min Presetting a lower limit of a range for the particle position; x is X max Presetting an upper limit of a range for the particle position; rand is arbitrarily at [0,1]And random numbers in the interval, wherein the first particle position data is any particle position data.
4. A method according to claim 3, wherein the DPF setting parameters comprise:
the exhaust gas volume flow PT1 filter coefficient, the flow resistance PT1 filter coefficient and the DPF inlet exhaust gas volume PT1 filter coefficient in the regeneration mode.
5. The method according to claim 1 or 2, wherein after selecting the best particle position data from all the obtained global optimal particle position data according to the fitness value corresponding to each global optimal particle, the method further comprises:
inputting the optimal particle position data and pre-acquired DPF experimental input data into a pre-constructed DPF simulation model to acquire DPF output simulation data;
and calculating the absolute difference between the output simulation data and DPF experimental output data by using an objective function.
6. The method of claim 5, wherein the DPF experimental input data comprises: the exhaust gas volume flow calculation value, the DPF surface temperature, the engine running state signal, the engine running time, the engine current running mode signal, the DPF pressure difference, the DPF function activation signal, the ash correction coefficient, the DPF carbon loading, the discontinuous DPF carbon loading simulation value, the DPF simulation state activation signal, the ash volume, the regeneration demand signal, the regeneration incomplete signal, the DPF simulation state activation state word, the environment temperature, the environment pressure, the pressure signal and the unreliable signal;
the DPF experimental output data comprises: carbon loading measured value, flow resistance after filtration in DPF, exhaust gas mass flow, carbon loading base value.
7. The method of claim 6, wherein the objective function is represented by equation 4:
wherein sim is a，b B-th discrete data contained in a-th simulation result data in the optimal particle position data; real a，b Represents the b-th discrete data contained in the a-th experimental output data in the DPF experimental output data, wherein sim a，b And real a，b There is a mapping relationship between them.
8. A DPF parameter tuning system, the system comprising: a processor and a memory;
the memory is used for storing one or more program instructions;
the processor being operative to execute one or more program instructions to perform the method of any one of claims 1-7.
9. A computer storage medium, characterized in that it contains one or more program instructions for performing the method according to any one of claims 1-7 by a DPF parameter tuning system.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911047575.8A CN110990941B (en) | 2019-10-30 | 2019-10-30 | DPF parameter setting method, system and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911047575.8A CN110990941B (en) | 2019-10-30 | 2019-10-30 | DPF parameter setting method, system and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110990941A CN110990941A (en) | 2020-04-10 |
| CN110990941B true CN110990941B (en) | 2023-07-21 |
Family
ID=70082792
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911047575.8A Active CN110990941B (en) | 2019-10-30 | 2019-10-30 | DPF parameter setting method, system and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110990941B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1901396A (en) * | 2006-07-27 | 2007-01-24 | 上海交通大学 | Method for forming multiple goal optimized array antenna direction pattern based on evolution algorithm |
| CN106941663A (en) * | 2017-05-16 | 2017-07-11 | 重庆邮电大学 | It is a kind of to merge convex optimization and the UWB localization methods of multi-objective particle swarm |
| CN108399450A (en) * | 2018-02-02 | 2018-08-14 | 武汉理工大学 | Improvement particle cluster algorithm based on biological evolution principle |
-
2019
- 2019-10-30 CN CN201911047575.8A patent/CN110990941B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1901396A (en) * | 2006-07-27 | 2007-01-24 | 上海交通大学 | Method for forming multiple goal optimized array antenna direction pattern based on evolution algorithm |
| CN106941663A (en) * | 2017-05-16 | 2017-07-11 | 重庆邮电大学 | It is a kind of to merge convex optimization and the UWB localization methods of multi-objective particle swarm |
| CN108399450A (en) * | 2018-02-02 | 2018-08-14 | 武汉理工大学 | Improvement particle cluster algorithm based on biological evolution principle |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110990941A (en) | 2020-04-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US7505949B2 (en) | Process model error correction method and system | |
| EP3018541B1 (en) | Configurable inferential sensor for vehicle control systems | |
| CN110121708B (en) | Process for designing a virtual sensor, corresponding virtual sensor, system and computer program product | |
| Gutjahr et al. | Advanced modeling and optimization for virtual calibration of internal combustion engines | |
| Jiang et al. | Implementation of model-based calibration for a gasoline engine | |
| Jin et al. | Benchmarks for model transformations and conformance checking | |
| CN113239963B (en) | Method, device, equipment, vehicle and storage medium for processing vehicle data | |
| CN110990941B (en) | DPF parameter setting method, system and storage medium | |
| CN113221271A (en) | Digital twin-driven quantitative recognition method for cracks of rotating blades of aircraft engine | |
| CN112597610A (en) | Optimization method, device and equipment for lightweight design of mechanical arm structure | |
| CN106126670B (en) | Operation data sorting processing method and device | |
| CN111177864B (en) | Particle swarm optimization-based internal combustion engine combustion model parameter optimization method and device | |
| CN114033533B (en) | DPF active regeneration period determination method and device, electronic equipment and storage medium | |
| ATE484661T1 (en) | METHOD FOR DETERMINING THE LOADING OF A PARTICLE FILTER | |
| CN115270932A (en) | Vehicle emission parameter optimization method and system, electronic device and storage medium | |
| CN110457863B (en) | Parameter estimation method for wind driven generator pitch subsystem based on ellipsoid contraction filtering | |
| CN116086560B (en) | Multi-factor coupled engine universal oil consumption correction method and system | |
| CN115906548A (en) | Automobile rim rigidity modeling method and device and storage medium | |
| CN114492245A (en) | Method and device for calculating and optimizing carbon accumulation in DPF (diesel particulate filter) and computer equipment | |
| CN116957423B (en) | Production environment monitoring and regulating method and system for packaging film production line | |
| CN113640115B (en) | Optimization method and system suitable for solving inverse problem of quasi-isentropic compression experimental data | |
| CN113434571B (en) | Knowledge mining method, platform, system and storage medium | |
| CN114896766B (en) | DPF filtering efficiency calibration method and device and electronic equipment | |
| CN115600397A (en) | DPF original exhaust carbon-loaded model calibration method and device, computer equipment and storage medium | |
| Gajare et al. | Calibration Tool for ECU based on CCP |
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 |