Search

CN-121993332-A - Intelligent control system and control method for starter

CN121993332ACN 121993332 ACN121993332 ACN 121993332ACN-121993332-A

Abstract

The invention relates to the technical field of starter control, and provides an intelligent starter control system and a control method. The signal sensing module acquires dynamic parameters such as starter current, crankshaft rotation speed and phase in real time; the environment information module acquires static and quasi-static parameters such as the temperature of the cooling liquid, the health state of the storage battery and the like; the intelligent decision unit is internally provided with a trained feedforward neural network model, the type of start failure can be diagnosed, the execution control module controls the starter to supply power according to the parameter set, and the method comprises the steps of parameter acquisition, initial control parameter generation, dynamic acquisition, real-time protection and restarting strategy execution after failure. The invention solves the problems of fixed control parameters, poor adaptability and low failed retry efficiency of the traditional starting system, and improves the starting reliability, the environmental adaptability and the service life of components.

Inventors

  • GU QIANHU
  • GU RENXI
  • ZHENG YANKAI
  • DAI GUOLIN

Assignees

  • 常州市松泽电器有限公司

Dates

Publication Date
20260508
Application Date
20260325

Claims (9)

  1. 1. An intelligent control system for a starter, comprising: The signal sensing module is used for collecting dynamic parameters in the starting process in real time and at least comprises starter current, engine crankshaft rotation speed and phase; The environment information module is used for acquiring static and quasi-static parameters before starting and at least comprises the temperature of engine cooling liquid, the estimated value of the health state of the storage battery and the standing time of the engine; the intelligent decision unit is configured to generate and output a dynamic optimal starting control parameter set in real time through the engine starting performance mapping model based on the dynamic parameters and static and quasi-static parameters, wherein the parameter set at least comprises an adaptive stall protection threshold parameter dynamically bound with a crankshaft phase; and the execution control module is controlled by the intelligent decision unit and is used for controlling the power supply of the starter according to the optimal starting control parameter set.
  2. 2. The intelligent starter control system according to claim 1, wherein the adaptive stall protection threshold parameter dynamically bound to crankshaft phase is used to determine according to the formula Calculating in real time the allowable current upper limit at the present phase Wherein, the method comprises the steps of, For the real-time crankshaft phase position, For the real-time rotation speed of the crankshaft, 、 、 As the parameters output by the engine starting performance map model, The function value of the phase interval near the compression top dead center of the engine is larger than the function value of the other phase intervals.
  3. 3. The intelligent starter control system according to claim 1, wherein the engine starting performance map model is a feedforward neural network model, the inputs of which are normalized eigenvectors composed of the dynamic parameters and static and quasi-static parameters, and the outputs of which at least include three-dimensional parameter vectors for calculating a stall protection threshold And a policy vector for characterizing different restart policy recommendation probabilities.
  4. 4. The intelligent control system of a starter of claim 1, wherein the intelligent decision unit is further configured to analyze a maximum crank angle recorded during a current start after a start failure; According to the preset angle interval of the maximum crank rotation angle, diagnosing the start failure type as one of serious clamping stagnation, partial rotation or approaching success; The generation of the restart strategy depends at least in part on the start failure type.
  5. 5. The starter intelligent control system of claim 4, further comprising an interface in communication with an engine controller; When the failed start type is diagnosed as "near successful", the intelligent decision unit sends a control parameter adjustment request to an engine controller via the interface, the request including at least an instruction to adjust the fuel injection amount or the ignition timing.
  6. 6. A starter intelligent control method using a starter intelligent control system according to any one of claims 1 to 5, comprising the steps of: s1, when a starting instruction is received, acquiring static and quasi-static parameters from an environment information module; S2, generating an initial optimal starting control parameter set through an engine starting performance mapping model based on the static and quasi-static parameters; s3, controlling the starter to start, and collecting dynamic parameters in real time in the starting process; S4, based on the dynamic parameters and the optimal starting control parameter set, executing real-time protection judgment and control; s5, if the starting fails, determining and executing a restarting strategy based on a failure analysis result and the engine starting performance mapping model.
  7. 7. The intelligent control method of a starter according to claim 6, wherein the step of executing the real-time protection judgment and control in S4 specifically includes: Parameters output according to the model 、 、 Crank phase acquired in real time And rotational speed By the formula Calculating the allowable current upper limit of the current phase; Comparing the starter current acquired in real time with the current upper limit; and if the starter current exceeds the upper current limit and the duration exceeds a preset threshold, determining that the starter current is at risk of stalling and cutting off starter power supply.
  8. 8. The intelligent control method of a starter according to claim 6, wherein the failure analysis result based on S5 specifically includes: Obtaining the maximum crank rotation angle when the starting fails; If the maximum crank rotation angle is smaller than the first threshold value, the analysis result is 'serious clamping stagnation', and the corresponding restarting strategy is to execute safe locking or try pulse mode starting after long delay; If the maximum crank rotation angle is larger than or equal to a first threshold value but smaller than a second threshold value, the analysis result is 'partial rotation', and the corresponding restarting strategy is to immediately try pulse mode starting; If the maximum crank rotation angle is larger than or equal to the second threshold value, the analysis result is that the engine is close to successful, and the corresponding restarting strategy is that the engine control parameter is adjusted and restarted after a short delay.
  9. 9. The intelligent control method according to claim 6, wherein the engine starting performance map model in S2 is trained offline by: collecting a plurality of process data samples started under different environmental conditions and vehicle states, each sample comprising an input feature vector and a corresponding labeling result; The input feature vector comprises normalized static and quasi-static parameters and dynamic parameters; for a successfully started sample, the labeling result is an optimal protection threshold parameter which is reversely deduced through an optimization algorithm; For a failed starting sample, the labeling result is an optimal recovery strategy label determined based on expert rules; and training the neural network model by using the sample set, so that the model learns the mapping from the input characteristic vector to the optimal protection threshold parameter and the recovery strategy.

Description

Intelligent control system and control method for starter Technical Field The invention relates to the technical field of starter control, in particular to an intelligent starter control system and a control method. Background At the moment of rapid development of the automobile industry, an engine starting system is taken as a vehicle core component, and the reliability and the intelligent level of the engine starting system directly influence the use experience and the running safety of a vehicle. The method has higher requirements on the adaptability and the accuracy of starting control whether the method is used in complex environments such as low temperature, high altitude and the like or in vehicle condition changes such as storage battery aging, engine abrasion and the like. The traditional starting system adopts a fixed parameter control mode, so that the dynamic adaptation requirements under different working conditions are difficult to meet, and the starting performance and stability are improved by intelligent technology upgrading. The existing starter control technology has the remarkable limitations that firstly, a stalling protection threshold is a fixed value, dynamic parameters such as a crank shaft phase and a rotating speed are not combined for dynamic adjustment, so that critical phases such as a compression top dead center are prone to be protected excessively or insufficiently, and further the burning or starting failure of a starter is caused, secondly, a unified retry mode is not adopted after the starting failure, an optimal scheme cannot be adapted for different failure types, the retry success rate is low, the component loss is possibly aggravated, thirdly, the control parameters are not fused with static parameters of a vehicle condition, the suitability is poor, and the starting efficiency is greatly reduced under an extreme environment or a vehicle aging scene. Disclosure of Invention Aiming at the defects of the prior art, the invention provides an intelligent control system and a control method for a starter, which solve the problems of low starting success rate and easy component loss under complex environment and vehicle conditions caused by the fact that the control parameters of the traditional starter control system are fixed, the starting failure diagnosis is inaccurate and the restarting strategy is lack of pertinence. In order to achieve the purpose, the invention is realized by the following technical scheme that the intelligent control system and the control method of the starter comprise the following steps: The intelligent decision unit is connected with the signal sensing module and the environment information module, is internally provided with a trained engine starting performance mapping model, and is configured to generate and output a dynamic optimal starting control parameter set in real time through the engine starting performance mapping model based on the dynamic parameters and the static and quasi-static parameters, wherein the parameter set at least comprises an adaptive stall protection threshold parameter dynamically bound with the crankshaft phase, and the execution control module is controlled by the intelligent decision unit and is used for controlling the power supply of the starter according to the optimal starting control parameter set. Preferably, the adaptive stall protection threshold parameter dynamically binding to crankshaft phase is used for determining the phase of the crankshaft according to the formulaCalculating in real time the allowable current upper limit at the present phaseWherein, the method comprises the steps of,For the real-time crankshaft phase position,For the real-time rotation speed of the crankshaft,、、As the parameters output by the engine starting performance map model,The function value of the phase interval near the compression top dead center of the engine is larger than the function value of the other phase intervals. Preferably, the engine starting performance mapping model is a feedforward neural network model, the input of the model is a normalized feature vector composed of the dynamic parameters and static and quasi-static parameters, and the output of the model at least comprises a three-dimensional parameter vector for calculating a stall protection thresholdAnd a policy vector for characterizing different restart policy recommendation probabilities. Preferably, the intelligent decision unit is further configured to analyze a maximum crank angle recorded during the current start after the start failure, diagnose the start failure type as one of a "serious stuck", "partial rotation" or "near successful" according to a preset angle interval to which the maximum crank angle belongs, and the generation of the restart strategy depends at least in part on the start failure type. Preferably, the system further comprises an interface in communication with the engine controller, and when the start failure type is diag