Remote Use Of A Linear Axis Rapid Development System

A Linear Axis Rapid Development System (RDS) was developed and tested in a previous research study. The Linear Axis RDS, which is based on Matlab Simulink, provides the student with a tool to explore all phases of controller development (i.e., simulation, emulation, and implementation) after the theoretical work is complete. However, the Linear Axis RDS did not provide the students with the ability to explore the linear axis dynamic model. In this paper, the previously developed Linear Axis RDS is augmented such that the student can model the linear axis dynamics and analyze their model. The student encodes their linear axis dynamic model as a subsystem in Matlab Simulink. The subsystem inputs and outputs, along with their engineering units, are carefully specified. The student then utilizes the Linear Axis RDS to analyze the dynamic model. The Linear Axis RDS has two modes: simulation and implementation. In the simulation mode the student simulates a linear axis system dynamic response for a variety of command voltages signals. The student can specify the magnitude and frequency of square, triangle, and sinusoidal command voltage signals, or they can create their own command voltage signal. In this mode the student can check their linear dynamic model for obvious errors (e.g., it is unstable). In the implementation mode, command voltage signals are simultaneously sent to the Matlab Simulink model and the real linear axis system. Simulation and experimental data are gathered and compared. This paper describes the modifications made to the Linear Axis RDS such that students can perform dynamic modeling and analysis of the linear axis. The results of an initial usability study are presented and analyzed. The Linear Axis RDS is then implemented remotely in an actuators course being taken by students in a Mechatronics program in the university ESIGELEC in Rouen, France. Introduction A Rapid Development System (RDS) for a Linear Axis was developed in [1]. A RDS is a software environment that allows students to rapidly integrate their controller and analyze it via simulation, emulation, and implementation. In the simulation mode the student simulates a linear axis system that includes their controller and detailed models of the interface hardware and linear axis. In the emulation mode, the simulation is performed on the computer hardware that will implement the controller. In this mode the student can ensure their algorithm will run in real time (i.e., the algorithm’s execution time is less than the sample period). In the implementation mode, the controller is deployed on the hardware system and experimental data is gathered. The Linear Axis RDS aided the students in the implementation of their controllers, even if the student had very little knowledge of control system hardware. The Linear Axis also allowed the students to focus on controller analysis. This paper adds a functionality to the Linear Axis RDS, known as the Axis Modeling function. This function allows students to develop and analyze a model of the Linear Axis. The Axis Modeling function was implemented in a short course on actuators given to eight students at ESIGELEC in Rouen, France. The students utilized the Axis Modeling function in the Linear P ge 15027.2 Axis RDS remotely. An implementation of this functionality is provided and the results are discussed in detail. Linear Axis Rapid Development System A Linear Axis RDS has been developed in [1] for linear axis controller design. This paper describes the addition of linear axis modeling capabilities to the Linear Axis RDS and the application to to the x–axis of the mini–CNC machine (Figure 1). A host computer, which has a non real–time operating system, runs the Linear Axis RDS Graphical User Interface (GUI), as well as the controller and models in simulation mode. A target computer, which has a real–time operating system, runs the controller and models in emulation and implementation modes. The target computer has a National Instruments (NI) 6711 digital to analog (D/A) output board and a NI 6602 counter–time (C/T) board. The target computer, which utilizes the xPC real–time operating system, sends voltage commands through the output board and receives encoder measurements through the C/T board, both at a sampling rate of 1 kHz. The output voltage range is ±10 V and is amplified before reaching the DC motor by a factor of 2.4. The motor has a gear ratio of 20/1 from the motor shaft to the output shaft. The output shaft is connected to a lead screw with a pitch of 5.9 mm/rev, which translates the linear axis. The encoder, which has a resolution of 500 counts/rev, measures the motor shaft displacement. The C/T board is run in quadrature mode for an effective encoder resolution of 2000 counts/rev. The Linear Axis RDS has two functions: Axis Controller and Axis Modeling. When the Axis Controller mode is utilized, the Linear Axis RDS allows the student to operate the controller in three different modes: simulation, emulation, and implementation. These modes are reviewed below. Then, the new Axis Modeling function is decribed. The Axis Controller function has a Simulation mode that allows the student to run the Simulink model of the closed–loop system on a host computer. This is beneficial because the student does not need to be connected to the physical linear axis. The Simulink model used for simulation mode, shown in Figure 2, contains four subsystems: Reference Generator, Controller, Computer– System Interface, and Linear Axis Model. The Reference Generator subsystem contains code that generates the desired linear axis position, called the reference position, which is sent to the controller. The Controller subsystem contains the controller to be tested. The Controller subsystem receives the reference and measured positions from the Reference Generator and Linear Axis Model subsystems, respectively, and sends the command voltage to the Computer– System Interface subsystem. The Computer–System Interface subsystem, shown in Figure 3, contains the simulated quantization and saturation affects of the D/A output board and the simulated quantization affect of the encoder. The Linear Axis Model subsystem, shown in Figure 4, contains the linear axis dynamic model. The model calculates the linear axis position given the command voltage from the Computer–System Interface subsystem. The reference position, measured position, and commanded voltage are recorded for controller performance analysis. The Axis Controller function has a Emulation mode that allows the student to run their controller on the target computer in real–time without moving the linear axis. The linear axis is simulated on the target computer. Emulation allows the student to determine if the target computer is able P ge 15027.3 to perform all computations and perform send and receive tasks within the specified sample period. The Simulink model used for the Emulation mode contains the same four blocks as the Simulink model used for Simulation mode. The difference between the Simulink models lies in the Computer–System Interface subsystem. The output voltage sent to the physical linear axis is zero and the position measurements from the encoder are disregarded as shown in Figure 5. These send and receive blocks in the Computer–System Interface subsystem are used to determine the time required to perform the communication tasks on the target computer. Since the linear axis is only one component of the mini–CNC, zero voltage signals are also sent to the other components to ensure they do not move. The same controller used in Simulation mode is also used in Emulation mode. The Axis Controller function has an Implementation mode that allows the student to operate the linear axis system using their controller. The Simulink model used for Implementation mode, shown in Figure 6, contains three subsystems: Reference Generator, Controller, and Computer– System Interface. The difference between the implementation and emulation modes is 1) there are no simulated affects in the Computer–System Interface subsystem and 2) there is no model to simulate the linear axis. This is because these affects are present in the computer interface hardware and physical linear axis and, thus, do not need to be simulated. Within the Computer– System Interface subsystem the command voltage from the controller is sent to the D/A output board and the encoder measurements are received from the C/T board as shown in Figure 7. The encoder measurements are used to calculate the measured linear axis position, which is sent to the controller. The same controller used in Simulation and Emulation modes is also used in the Implementation mode. The new function that has been added to the Linear Axis RDS is the Axis Modeling function. When used for axis modeling, the Linear Axis RDS can be used to collect system dynamic response data for modeling, simulate the linear axis dynamic response, and compare the model dynamic response to measured data. Three modes are provided for the student to perform axis modeling: Axis Test, Simulation, and Model Validation. The Axis Test mode allows the student to collect the linear axis dynamic response data for different command voltage signals. This data can be used to model the axis. The Axis Test mode Simulink model is shown in Figure 8. The command voltage is generated in the Voltage Signal subsystem and sent to the physical system via the IO Interface subsystem. The command voltage and measured position are recorded for further analysis. The Simulation mode allows the student to run the Simulink model on the host computer with a sine wave or a user defined command voltage signal. The Simulation mode Simulink model is shown in Figure 9. The command voltage generated in the Voltage Signal subsystem is sent to the Component model via the Computer–System Interface subsystem, which simulates the effects of physical input/output system such as quantizatio