Estimation of Intake Manifold Absolute Pressure Using Kalman Filter

For vehicles with intake manifold absolute pressure (MAP) sensor, the intake air mass is obtained using speed-density method. Since the analog MAP signal will contain high frequency noise with uncertain amplitude, the MAP value obtained in the engine management system using angle based sampling will result in MAP value variation even for engine steady-state operation. In order to properly obtain a MAP value under nonlinear time-varying characteristics, a MAP estimation method based on a closed-loop model is proposed. First, an adaptive two-input single-output intake manifold model is constructed. The Recursive Least Square technique is utilized to on-line identify the intake manifold model with throttle opening angle and engine speed as inputs. The identified intake manifold model is then employed to estimate the MAP using the Kalman Filter. Simulation results show that the proposed method can bring smaller standard deviation of air fuel ratio than that of using conventional methods for noise rejection under open-loop fuel control and system parameters drift. When a high frequency noise with higher amplitude is caught while sampling a MAP value, the proposed method can also reduce the noise effect and preserve the open-loop control performance on air fuel ratio. The proposed method is also investigated if the engine output torque is fluctuated.