Using Ion Sensing with O2 Sensor Based Adaptive Calibration to Perform Cold Start Closed Loop A/F Ratio Control to Reduce Engine Cold Start HC Emissions, 03-9305Printer Friendly Version
Inclusive Dates: 04/01/02 - 10/01/03
Background - Currently, most gasoline engines are equipped with oxygen sensors to provide closed loop control of the air/fuel ratio. However, oxygen sensors have limitations. These sensors require approximately 30 seconds to warm up before accurate readings can be provided. The oxygen sensor is useless until it warms up, and 60 to 80 percent of total tailpipe hydrocarbon emissions are produced in the first 60 to 120 seconds after an engine cold start. During this interval, the air/fuel ratio is under open loop control and is not optimized. However, an ion-sensing spark plug, which detects ion signals, does not need this warm-up interval and can provide in-cylinder air/fuel ratio control as soon as the engine starts. Unlike other kinds of sensors, engine spark plugs are routinely replaced during periodic maintenance. In addition, production engines have many sources of variation and aging effects, so the technique of ion sensing for air/fuel ratio feedback requires online adaptive calibration to be successful.
Approach - This project aimed to use a signal from a warmed oxygen sensor to calibrate ion-sensing spark plugs (ion sensor) adaptively. The goal of this project is to minimize over-fueling the engine during the warm-up transient of the federal test protocol driving cycle. The research team investigated the capability of using ion sensing to provide air/fuel ratio information that is fast enough to allow the electronic control unit to perform closed loop control during cold starts.
Accomplishments - Five objectives have been
accomplished in this project.
Second, engine load, speed and coolant temperature effects were investigated on the ionization with different A/F ratio.
Third, the engine control system hardware and signal analyzing software were developed.
Fourth, correlation was established from both the constant volume combustion bomb investigation and engine test results and a neural network model for approximating the A/F ratio was developed.
Fifth, the online calibration mode has been established based on the engine test results.
The limitations of using in-cylinder ionization were discovered. The best mode of using in-cylinder ionization is between light and medium load and speed. Under high speed and load condition, the characteristics of the ionization will become different from that under lower load and speed and is difficult to use.
A utility patent application has been filed based on this idea.