Visteon simplifies automotive powertrain control using NI LabVIEW Control Design and Simulation Module

作者：Arek Dutka - Industrial Systems and Control Limited

　　Gustav Ferrao - Industrial Systems and Control Limited

　　“LabVIEW has two main advantages in today’s software used in the automotive industry: one is its front panel, which can serve as a powerful user interface; the other is its vivid development environment, which can avoid low-level language programming.”

　　challenge:

　　Validate complex automotive engine designs by simulating multiple variables to achieve optimal fuel consumption, engine performance, and emissions control.

　　Solution:

　　Using the NI LabVIEW Control Design and Simulation Module, we developed an application that can perform real-time control, analysis, and testing.

　　Today, automotive powertrain control systems must continue to evolve to meet demands that include regulating exhaust emissions to meet increasingly stringent emissions standards; providing better fuel efficiency to comply with corporate average fuel consumption standards; and meeting user needs for performance and comfort.

　　这些要求是相互联系的，甚至经常是相互矛盾的。比如，贫燃技术可以显著地减少油耗，但同时降低了三元催化转换的效率，造成了额外的空气污染。

　　There are two ways to meet today's automotive specifications: one is to improve existing structures, and the other is to introduce new and more complex mechanical designs.

　　Among the parameters that determine engine performance, camshaft shape is the most important design parameter.

　　During the design process, some engines focus on meeting torque requirements, while others focus on optimizing speed, so no single shape can meet the requirements of all design parameters.

　　There are four main variable cam timing strategies for dual overhead camshaft (DOHC) engines:

　　• Intake cam phase shift only (intake only)

　　• Exhaust cam phase shift only (exhaust only)

　　• Intake and exhaust cams are equally phased (both are equal)

　　• Independent phase shifting of intake and exhaust cams (dual independent)

　　In a twin-independent variable camshaft timing (TIVCT) engine, the intake and exhaust camshafts are independently calibrated as a function of valve position and engine speed.

　　The system offers a lot of freedom to improve engine performance, so it is necessary to find a way to optimize the valve timing parameters to obtain the best fuel consumption, engine performance and emission control.

　　However, the result of this technology is a highly complex real-time control algorithm. Although TIVCT has been introduced in the field of automotive engines several years ago, it is still the focus of research and exploration today.

　　Real-time control, analysis, and testing using LabVIEW

　　This project is based on TIVCT engine modeling and optimal controller design to achieve special engine performance requirements. The goal of the control strategy is to provide the engine with a reference amount of torque tracking, while minimizing the fuel consumption rate during braking and optimizing the stability of fuel combustion.

　　This project was completed using the LabVIEW Control Design and Simulation Module and its built-in linear algebra functions. Among the software used in the automotive industry today, LabVIEW has two main advantages: one is its front panel, which can serve as a powerful user interface; the other is its vivid development environment, which can avoid low-level language programming.

　　In addition, many NI hardware products integrate tools for control, design, and simulation to facilitate the development of real-time control, analysis, and testing applications, which also makes LabVIEW very attractive to automotive R&D departments.

　　For the engine model, the most important variables for control system operation include the air flow in the intake manifold, the position of the independent camshaft at the inlet and the exhaust valve exhaust timing relative to the crankshaft.

　　The control outputs are engine torque, brake fuel consumption rate and coefficient of variation of mean effective pressure indication. Other variables that affect system performance (such as engine speed and engine coolant temperature) are treated as external parameters and used as scheduling variables of the control system.

　　Using the LabVIEW Control Design and Simulation Module, the time-continuous TIVCT engine model combines a static characteristic equation of a typical combustion process with differential equations describing the actuator and intake manifold to create a dynamic model.

　　The resulting nonlinear TIVCT engine model has the characteristics of multiple input, multiple output (MIMO). By manipulating each input variable, its input-output relationship has a significant cross-effect. In this control application, LabVIEW is used to set the system to a specific operating point, linearize the nonlinear engine model, and thus develop a linear model.

　　Interactive simulation using LabVIEW front panels

　　An advanced optimization controller is designed using the linear quadratic regulator (LQR) in LabVIEW. Functionally, this controller achieves two goals: minimizing the offset and realizing the role of a calibrator. In the presence of external disturbances, the steady-state error can be eliminated by introducing an intra-loop integral, thereby achieving the above controller design goals.

　　In order to define the performance index and minimize the output error and output change rate, LabVIEW is used to perform state feedback and reference point tracking on the TIVCT engine based on the optimization theory of continuous time domain systems, and the expected gain is obtained through this tool.

　　The local controller and linear model were built and simulated in LabVIEW. The system tracks the engine torque with an accurate steady-state value related to the setpoint while minimizing the brake specific fuel consumption (BSFC) and the coefficient of variation of mean indicated pressure (COVIMEP).

　　Placing the two tuning variables Q and R on the front panel ensures intuitive detection of the controller and online adjustment, which also makes full use of the interactive simulation features of LabVIEW.

　　In order to easily transfer the simulation to computer hardware for the final application, the model and controller are usually implemented in discrete-time systems. Discrete-time controllers can be derived from continuous-time controllers or redesigned directly in discrete-time systems using the same Linear Quadratic Regulator VI.

　　Because the model is nonlinear, the ideal gain parameters that produce the desired response at one operating point may not produce the same satisfactory response at another operating point.

　　Therefore, it is necessary to use the corresponding ideal gain parameters in different working ranges of the nonlinear model to achieve gain scheduling. The interactive adjustment of parameters is completed through the front panel to rationalize the gain tuning process.

　　Use LabVIEW for interactive simulation, real-time control, analysis, and testing

　　Screen capture demonstrating multiple-input multiple-output (MIMO) design methodology