CN105224730A - The original NO of a kind of high pressure common rail electric-controlled diesel engine 2forecasting of discharged quantity method - Google Patents
The original NO of a kind of high pressure common rail electric-controlled diesel engine 2forecasting of discharged quantity method Download PDFInfo
- Publication number
- CN105224730A CN105224730A CN201510591913.XA CN201510591913A CN105224730A CN 105224730 A CN105224730 A CN 105224730A CN 201510591913 A CN201510591913 A CN 201510591913A CN 105224730 A CN105224730 A CN 105224730A
- Authority
- CN
- China
- Prior art keywords
- diesel engine
- gravity
- original
- model
- burning center
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Abstract
The invention discloses the original NO of Design of High Pressure Common Rail Diesel Engine under a kind of Accurate Prediction current working
2the method of discharge, belong to Exhaust Control for Diesel Engine field, the method comprises: obtain Key experiments data; With diesel engine speed, distributive value, injection advance angle, admission pressure, intake air temperature, oily rail pressure for input quantity, set up burning center of gravity by Local Linear Model tree algorithm and estimate submodel; Using diesel engine speed, air-fuel ratio, burning center of gravity as input quantity, set up NO by Local Linear Model tree algorithm
2forecasting of discharged quantity model; Online Real-time Obtaining diesel engine ECU data, input NO
2forecasting of discharged quantity model, can obtain the original NO of high pressure common rail electric-controlled diesel engine
2discharge capacity.The present invention can quick and precisely prediction of Diesel Engine NO
2original emission, is convenient to the original NO of diesel engine
2the monitoring of discharge and the control of SCR catalyst.
Description
Technical field
The present invention relates to the original NO of high pressure common rail electric-controlled diesel engine
2the Forecasting Methodology of discharge capacity, belongs to Exhaust Control for Diesel Engine field.
Background technology
Oxides of nitrogen (NitrogenOxides, NOx) is one of primary discharge pollutant of diesel engine, comprises nitrogen monoxide (NitricOxide, NO) and nitrogen dioxide (NitrogenDioxide, NO
2).NO
2be a kind of rufous high activity gas, have strong impulse effect to respiratory tract, its toxicity much larger than NO, short-term NO
2expose and have certain relation with the incidence of disease of breathing problem and cardiopulmonary relevant disease.NO simultaneously
2still produce acid rain, form one of photo-chemical smog and the principal element causing climate change, to air ecologic environment, there is great harm.NO
2be one of important indicator evaluating ambient air quality, " ambient air quality " (GB3095-1996) that the NAAQS that Environmental Protection Agency EPA promulgates, european union directive 1999/30/EC and China promulgate all defines the NO in surrounding air
2concentration limit.Therefore, while the discharge of control Diesel NOx, its NO is monitored
2discharge capacity is also extremely important.
At present, SCR technology SCR is the main aftertreatment technologies controlling Diesel NOx discharge, the original NO of diesel engine
2the NOx conversion efficiency of ratio to SCR catalyst that discharge accounts for NOx emission has material impact, therefore monitors the original NO of diesel engine
2discharge has vital role to SCR control technology.Be not directly used at present and measure diesel engine NO
2the physical sensors of discharge capacity, estimates original NO after measuring NOx discharge only with the NOx sensor of two by installing before and after SCR
2discharge.
Summary of the invention
The object of the invention is to overcome above-mentioned the deficiencies in the prior art, and provides a kind of high pressure common rail electric-controlled diesel engine original NO
2the Forecasting Methodology of discharge capacity, the present invention can quick and precisely prediction of Diesel Engine NO
2original emission, makes up the NO not used for diesel engine
2the deficiency of physical sensors, is convenient to the original NO of diesel engine
2the monitoring of discharge and the control of SCR catalyst.
The technical scheme realizing the object of the invention employing is: a kind of for the original NO of high pressure common rail electric-controlled diesel engine
2the Forecasting Methodology of discharge capacity, comprises the following steps:
(1) tested by high pressure common rail electric-controlled diesel engine bench, obtain and comprise diesel engine ECU operational factor, burning center of gravity and NO
2discharge capacity data;
(2) using diesel engine speed, distributive value, injection advance angle, admission pressure, intake air temperature, oily rail pressure 6 parameters as input quantity, burning center of gravity is as output quantity, based on Local Linear Model tree algorithm Modling model, and with the test figure obtained in step (1), model is checked, obtain burning center of gravity and estimate submodel;
(3) using diesel engine speed, air-fuel ratio, burning center of gravity as input quantity, diesel engine NO
2discharge capacity, as output quantity, based on Local Linear Model tree algorithm Modling model, and is checked model with the test figure obtained in step (1), is obtained NO
2forecasting of discharged quantity model;
(4) in diesel engine operational process, Real-time Obtaining diesel engine ECU data, the burning center of gravity obtained in diesel engine speed, distributive value, injection advance angle, admission pressure, intake air temperature, oily rail pressure 6 parameters input steps (2) is estimated submodel, thus obtains burning center of gravity estimated value;
(5) NO will obtained in burning center of gravity estimated value input step (3) obtained in diesel engine speed, air-fuel ratio and step (4)
2forecasting of discharged quantity model, can obtain diesel engine NO
2original emission.
The inventive method is applicable to high pressure common rail electric-controlled diesel engine, by electronic controlled diesel ECU service data, in conjunction with the combustion model of diesel engine, estimates diesel combustion center of gravity.Due to the original NO of diesel engine
2discharge has stronger correlativity, the burning center of gravity therefore will estimated with burning center of gravity, and diesel engine speed and air-fuel ratio are as NO
2the input quantity of forecasting of discharged quantity model, can calculate the original NO of diesel engine accurately
2discharge capacity.
Burning center of gravity of the present invention estimates submodel and NO
2forecasting of discharged quantity model is all set up based on Local Linear Model tree algorithm, and weighting function is normalized Gaussian function, and the parameter transmitting ride comfort after each segmentation is set to 0.35, and the threshold value that each segmentation global error reduces is set to 4.5%.Compared with BP neural network model, it is high that Local Linear Model tree has efficiency of algorithm, and the feature that computing is fast, achieves the NO of fast and reliable
2emitted smoke.
Accompanying drawing explanation
Fig. 1 is Local Linear Model tree algorithm process flow diagram;
Fig. 2 is the process flow diagram of the inventive method;
Fig. 3 is the Local Linear Model tree-model for estimating diesel combustion center of gravity;
Fig. 4 is the original NO of prediction of Diesel Engine
2the Local Linear Model tree-model of discharge capacity;
Fig. 5 is the original NO that the present invention predicts
2the Comparative result figure of discharge capacity and experimental test value.
Embodiment
Below in conjunction with the drawings and specific embodiments, the present invention is described in further detail.
Diesel NOx discharge is the N in the air in diesel engine suction cylinder
2and O
2reaction product at high temperature, mainly NO and NO
2.In the original NOx emission of diesel engine, NO
210 ~ 30% of total NOx in exhaust can be accounted for.It is generally acknowledged NO
2by flame zone NO and HO
2reaction generates, NO+HO
2 nO
2+ OH, then NO
2again NO is changed into again, i.e. NO by following reaction equation
2+ O
nO+O
2.As the NO in flame
2when mixing with cold fluid, its changing again to NO can be suppressed, thus quenching and remaining, directly discharge from tail gas.
NO
2the N in air
2and O
2the product reacted in cylinder of diesel engine, the combustion case in cylinder determines NO
2growing amount and generating rate, simultaneously also determine the combustion characteristic parameters such as maximum outbreak pressure, maximum combustion temperature, Pressure Rise Rate and burning center of gravity.Burning center of gravity is the crank angle of cylinder fuel combustion 50% correspondence, is the important parameter characterizing Combustion Characteristics of Diesel Engine.The present invention studies and finds the original NO of diesel engine
2discharge has stronger correlativity with burning center of gravity.
High pressure common rail electric-controlled diesel engine adopts Modem electronic control technical controlling cylinder of diesel engine combustion, the multiple sensors of installing can monitor diesel engine operational factor, diesel engine ECU can gather these sensor signals and calculate, and controls actuator operation, thus controls diesel engine combustion.
Local Linear Model tree is a kind of Fast Fuzzy neural network algorithm, the input layer of network is made up of primary data, the second layer is hidden layer, third layer is output layer, conversion from input layer space to hidden layer space is nonlinear, conversion from hidden layer space to output layer space is linear, and its work space is divided into M sub spaces by binary tree algorithm.The training of Local Linear Model tree is corresponding with its design feature, adopts the method for binary tree to carry out, and training is divided into inside and outside two-layer, and outer training optimizes its structure, internal layer training parameter.Outer field training comprises the division work space determining neuronic number, the training of structure is performed such, the number of first fixing hidden node is constant, the weight coefficient that only adjustment is relevant, then according to binary tree algorithm, the input space is repartitioned, therefrom select the bifurcated making training error minimum, as new hidden node, carry out successively, until select best structure.Concrete training process flow diagram as shown in Figure 1.
The present invention is based on the above-mentioned fact, by obtaining the operational factor of diesel engine ECU monitoring, in conjunction with trained Local Linear Model tree-model, estimating cylinder combustion and burning situation, and then estimating NO
2discharge capacity.
Flow process as shown in Figure 2, the present invention is for measuring the original NO of high pressure common rail electric-controlled diesel engine
2the method of discharge, specifically comprises the following steps:
High pressure common rail electric-controlled diesel engine test stand carries out universal characteristic test, tests the different operating points of diesel engine, operating mode is counted and is no less than 70.Diesel engine ECU data are read by CAN; Diesel combustion center of gravity parameter is obtained by cylinder pressure sensor and Combustion tester; The original NO of diesel engine is obtained by emission test equipment
2discharge capacity.
The present embodiment diesel combustion center of gravity used is estimated submodel and is realized by Local Linear Model tree as shown in Figure 3.Local Linear Model must be trained by the related data with known output or result by tree, therefore needs to use above-mentioned test figure.Train the input amendment collection parameter needed for Local Linear Model tree to have 6 in the present embodiment, comprise diesel engine speed, distributive value, injection advance angle, admission pressure, intake air temperature, oily rail pressure, output sample collection parameter is burning center of gravity.Each sample set data for model training are no less than 50.Test diesel combustion center of gravity being estimated to submodel shows: can realize convergence when segmentation times is no less than 30 times, reach higher training precision.Weighting function is normalized Gaussian function, and the parameter transmitting ride comfort after each segmentation is set to 0.35, and the threshold value that each segmentation global error reduces is set to 4.5%.
The original NO of the present embodiment diesel engine used
2forecasting of discharged quantity model is realized by Local Linear Model tree as shown in Figure 4.Local Linear Model must be trained by the related data with known output or result by tree, therefore needs to use above-mentioned test figure.Train the input amendment collection parameter needed for Local Linear Model tree to have 3 in the present embodiment, comprise diesel engine speed, burning center of gravity and air-fuel ratio, output sample integrates parameter as NO
2discharge capacity.Each sample set data for model training are no less than 50.To the original NO of diesel engine
2the test of forecasting of discharged quantity model shows: can realize convergence when segmentation times is no less than 22 times, reach higher training precision.Weighting function is normalized Gaussian function, and the parameter transmitting ride comfort after each segmentation is set to 0.35, and the threshold value that each segmentation global error reduces is set to 4.5%.
The diesel combustion center of gravity trained is estimated submodel and original NO
2forecasting of discharged quantity model is implanted in high pressure common rail electric-controlled diesel engine ECU, in diesel engine operational process, by gathering ECU signal, operational parameter value input burning center of gravity is estimated submodel and original NO
2forecasting of discharged quantity model, can obtain the original NO of high pressure common rail electric-controlled diesel engine
2forecasting of discharged quantity value.
Use the original NO of prediction of Diesel Engine of the present invention
2as shown in Figure 5, what input under each data point in figure illustrates different operating point that training parameter produces predicts the outcome and test result actual under corresponding operating mode the results contrast of concentration of emission and test measurements.Local Linear Model tree, after training, just can carry out new test with checking when not having Output rusults, and whether input training parameter can produce that precision is sufficiently high to predict the outcome again.After analyzing predicting the outcome, the deviation of display predict and actual result is within rational level.
The inventive method to the weighting function of Local Linear Model tree algorithm, transmit ride comfort parameter and global error and reduce threshold value and be optimized, the diesel combustion center of gravity set up based on this algorithm estimates submodel and original NO
2forecasting of discharged quantity model needs input parameter few, and efficiency of algorithm is high, fast operation, and implantable diesel engine ECU carries out real-time online operation.
The input of the inventive method data used depends on the existing sensor of high pressure common rail electric-controlled diesel engine and exports, and does not need to increase extra cylinder pressure sensor, reduces costs.
Claims (2)
1. one kind for the original NO of high pressure common rail electric-controlled diesel engine
2the Forecasting Methodology of discharge capacity, is characterized in that, comprises the following steps:
(1) tested by high pressure common rail electric-controlled diesel engine bench, obtain diesel engine ECU operational factor, burning center of gravity and NO
2discharge capacity data;
(2) using diesel engine speed, distributive value, injection advance angle, admission pressure, intake air temperature, oily rail pressure 6 parameters as input quantity, burning center of gravity is as output quantity, based on Local Linear Model tree algorithm Modling model, and with the test figure obtained in step (1), model is checked, obtain burning center of gravity and estimate submodel;
(3) using diesel engine speed, air-fuel ratio, burning center of gravity as input quantity, diesel engine NO
2discharge capacity, as output quantity, based on Local Linear Model tree algorithm Modling model, and is checked model with the test figure obtained in step (1), is obtained NO
2forecasting of discharged quantity model;
(4) in diesel engine operational process, Real-time Obtaining diesel engine ECU data, the burning center of gravity obtained in diesel engine speed, distributive value, injection advance angle, admission pressure, intake air temperature, oily rail pressure 6 parameters input steps (2) is estimated submodel, thus obtains burning center of gravity estimated value;
(5) NO will obtained in burning center of gravity estimated value input step (3) obtained in diesel engine speed, air-fuel ratio and step (4)
2forecasting of discharged quantity model, can obtain diesel engine NO
2original emission.
2. one according to claim 1 is used for the original NO of high pressure common rail electric-controlled diesel engine
2the Forecasting Methodology of discharge capacity, is characterized in that: the burning center of gravity in step (2) estimates the NO in submodel and step (3)
2forecasting of discharged quantity model is all set up based on Local Linear Model tree algorithm, and the weighting function of this model is normalized Gaussian function, and the parameter transmitting ride comfort after each segmentation is 0.35, and the threshold value that each segmentation global error reduces is 4.5%.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510591913.XA CN105224730A (en) | 2015-09-17 | 2015-09-17 | The original NO of a kind of high pressure common rail electric-controlled diesel engine 2forecasting of discharged quantity method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510591913.XA CN105224730A (en) | 2015-09-17 | 2015-09-17 | The original NO of a kind of high pressure common rail electric-controlled diesel engine 2forecasting of discharged quantity method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105224730A true CN105224730A (en) | 2016-01-06 |
Family
ID=54993696
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510591913.XA Pending CN105224730A (en) | 2015-09-17 | 2015-09-17 | The original NO of a kind of high pressure common rail electric-controlled diesel engine 2forecasting of discharged quantity method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105224730A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106991507A (en) * | 2017-05-19 | 2017-07-28 | 杭州意能电力技术有限公司 | A kind of SCR inlet NOx concentration on-line prediction method and device |
| CN109492807A (en) * | 2018-11-01 | 2019-03-19 | 大唐环境产业集团股份有限公司 | Based on the boiler NO for improving quanta particle swarm optimizationXPrediction model optimization method |
| CN109944706A (en) * | 2019-03-29 | 2019-06-28 | 潍柴动力股份有限公司 | The regulation method and device of oxynitride discharge |
| CN113611375A (en) * | 2021-08-09 | 2021-11-05 | 成都佳华物链云科技有限公司 | Data determination method, device, equipment and storage medium in thermal power plant system |
| CN117854636B (en) * | 2024-03-07 | 2024-04-30 | 西南林业大学 | Method for predicting emission quantity of particulate matters in transient process of diesel vehicle |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103410592A (en) * | 2013-07-18 | 2013-11-27 | 武汉理工大学 | Diesel NOx original emission load predicting method based on crankshaft angular velocity sensor |
| CN104715142A (en) * | 2015-02-06 | 2015-06-17 | 东南大学 | NOx emission dynamic soft-sensing method for power station boiler |
-
2015
- 2015-09-17 CN CN201510591913.XA patent/CN105224730A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103410592A (en) * | 2013-07-18 | 2013-11-27 | 武汉理工大学 | Diesel NOx original emission load predicting method based on crankshaft angular velocity sensor |
| CN104715142A (en) * | 2015-02-06 | 2015-06-17 | 东南大学 | NOx emission dynamic soft-sensing method for power station boiler |
Non-Patent Citations (2)
| Title |
|---|
| 何超等: "高压共轨柴油机燃烧与二氧化氮排放特性研究", 《内燃机工程》 * |
| 王新宇: "基于局部线性模型树的高压共轨柴油机排放模型", 《车用发动机》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106991507A (en) * | 2017-05-19 | 2017-07-28 | 杭州意能电力技术有限公司 | A kind of SCR inlet NOx concentration on-line prediction method and device |
| CN109492807A (en) * | 2018-11-01 | 2019-03-19 | 大唐环境产业集团股份有限公司 | Based on the boiler NO for improving quanta particle swarm optimizationXPrediction model optimization method |
| CN109944706A (en) * | 2019-03-29 | 2019-06-28 | 潍柴动力股份有限公司 | The regulation method and device of oxynitride discharge |
| CN109944706B (en) * | 2019-03-29 | 2022-06-24 | 潍柴动力股份有限公司 | Method and device for regulating and controlling nitrogen oxide emission |
| CN113611375A (en) * | 2021-08-09 | 2021-11-05 | 成都佳华物链云科技有限公司 | Data determination method, device, equipment and storage medium in thermal power plant system |
| CN113611375B (en) * | 2021-08-09 | 2023-10-10 | 成都佳华物链云科技有限公司 | Method, device, equipment and storage medium for determining data in thermal power plant system |
| CN117854636B (en) * | 2024-03-07 | 2024-04-30 | 西南林业大学 | Method for predicting emission quantity of particulate matters in transient process of diesel vehicle |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8478565B2 (en) | Method of monitoring soot mass in a particulate filter and monitoring system for same with correction for active regeneration inefficiency | |
| US8365586B2 (en) | Method of monitoring soot mass in a particulate filter and monitoring system for same | |
| CN104204481B (en) | Method and apparatus for monitoring gas sensor | |
| US7299123B2 (en) | Method and device for estimating the inlet air flow in a combustion chamber of a cylinder of an internal combustion engine | |
| CN102216573B (en) | Controller of internal combustion engine, and device for measuring mass flow of nox refluxed back to intake passage along with blow-by gas | |
| AU2010334853B2 (en) | Method and apparatus for measuring and controlling the EGR rate in a combustion engine. | |
| RU2569937C2 (en) | Method and device for control over ice moisture content transducer including measurement of oxygen concentration by other engine transducers such as nitrogen oxide transducer, lambda-probe and/or oxygen transducer | |
| CN105224730A (en) | The original NO of a kind of high pressure common rail electric-controlled diesel engine 2forecasting of discharged quantity method | |
| US11199120B2 (en) | Inferential flow sensor | |
| CN103410592B (en) | Diesel NOx original emission load predicting method based on crankshaft angular velocity sensor | |
| US20140012791A1 (en) | Systems and methods for sensor error detection and compensation | |
| US20150190749A1 (en) | Method and Apparatus for Estimating the Amount of Reductant Slip in a Selective Catalytic Reduction Device | |
| US10620174B2 (en) | Method for improving accuracy of sensor outputs for measuring NOx, and system and apparatus for same | |
| GB2461301A (en) | A method for detecting faults in the air system of internal combustion engines | |
| JP2011099666A5 (en) | ||
| CN110827444B (en) | Heavy vehicle emission factor obtaining method suitable for OBD remote emission monitoring data | |
| CN107023367B (en) | A kind of SCR system of diesel engine ammonia input pickup fault diagnosis and fault tolerant control method | |
| CN113884307A (en) | Method and system for detecting accuracy of air inflow sensor | |
| US11561212B1 (en) | Method for determining NOx sensor data falsification based on remote emission monitoring | |
| CN114622974B (en) | Intelligent detection and diagnosis system and method for motor vehicle exhaust | |
| CN112177738B (en) | Urea consumption monitoring method and diesel engine | |
| CN116804387A (en) | System and method for estimating emissions | |
| CN117848578A (en) | Differential pressure signal detection method, differential pressure signal detection equipment and differential pressure signal detection medium | |
| CN115754143A (en) | Excavator emission test system and method and excavator | |
| Matušů et al. | Mathematical model of the emissions of a selected vehicle |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160106 |
|
| RJ01 | Rejection of invention patent application after publication |