Real-time computing for a mechatronic system

The real-time target computer subsystem

The real-time target computer subsystem controls the electromechanical subsystem. This control will be able to answer questions such as, “How much current should be applied to the motor right now?” The “right now” refers to the real-timeness of the computing. A real-time computer’s latency is the duration between the moment of an input change, such as a measurement or a user command, and an output change, such as the current applied to a motor. In control systems, the required latency is related to the speed at which the system under control (known as the “plant”) changes. A quickly changing plant requires a real-time computer with smaller latencies than those of a slowly changing plant.

Many mechanical systems of interest respond rather quickly, in milliseconds to seconds. The constraints on the computing system, then, are rather rigorous. Real time waits for no computer. Therefore, timing considerations must be paramount in our consideration of the computing system, and we must treat the computer as integral to the entire system.

How does the computer determine what should be done? Answers to this question fall into three categories:

User commands

The user of the system provides the system with commands about its behavior, such as “spin at \(300\) revolutions per minute (RPM).”

Programming

The engineer programs the computer to interpret the user’s commands and instantiate a control algorithm. This control algorithm computes control signals that interact with the electronic system and thereby influence the mechanical system’s behavior.

Feedback

In many cases, the state of the mechatronic system is measured with sensors (e.g., encoders), providing the computer with feedback about the system, which the control algorithm uses to compute control signals.

As we explore the other subsystems of the target system, we will encounter the user commands and feedback. For programming, we will have to wait for further discussion of the development system.

The target computer

The target computer is the . This is the computer that controls the target system via the C programs that we will write and the user input.

The myRIO board, pictured in figure 0.2, is designed around a Xilinx Z-7010 system on a chip (SoC), which includes both an processor for general programming and a field-programmable gate array (FPGA) for fast input/output (I/O), which we will use for, among other I/O tasks, counting the encoder’s rotation.

The myRIO runs a version of Real-Time Linux, an RTOS.² It can run programs in a variety of programming languages, but we will program it in C, which is the language of choice for most real-time programming.³ In each lab exercise, we will write and execute C programs on the myRIO.

The myRIO is described in detail in appendix A. If you are building your own target system, please refer to that appendix.

The user-interface subsystem

The user of a target system may be, for example, a person using a consumer device, a worker in a manufacturing plant tasked with adjusting automation settings, or an engineer programming a robot for complex tasks. The user must be able to input high-level commands that prescribe the behavior of the device. The general term for this interface is user input device. This is often an array of mechanical buttons, a keypad, or a hardware keyboard, as illustrated in figure 0.3. In the T1 target system, we will use a keypad and momentary contact buttons.

In most cases, the user must also be able to receive information from the target computer. Such information can include, for example, menu options, system status, and measured system variables. In most cases, this is some sort of display, although it can be as simple as a light-emitting diode (LED) indicator. Displays can be simple \(n\)-segment displays, dot-matrix liquid crystal displays (LCDs), or LCD/LED screens, as illustrated in figure 0.4.

The T1 target system user interface devices are a keypad and a display. The keypad is a 16-key in a \(4 \times 4\) arrangement, depicting the \(10\) decimal digits and some control characters, as shown in . The display is a four-row dot-matrix display with 20 characters per row, as shown in .

In lab 1, lab 2, lab 3, we will write programs for the myRIO to communicate with these devices. In later lab exercises, we will continue to solicit user input through the keypad and present information on the display.

Appendix A describes the two UI devices in detail.

The electromechanical subsystem

As shown in figure 0.1, the T1 target electromechanical subsystem, under the control of the target computer, supplies power to a motor, which rotates a flywheel (inertial) load. To drive the motor, a variable amount of electrical power will be required. A power supply supplies electrical power to the circuit. In general, this can be alternating current (AC) or direct current (DC), but the T1 target system will require a DC power supply. However, a DC power supply is not adjustable by itself—some fixed voltage or current will be supplied to any electrical load connected, within power limitations.

How can the target computer’s control decisions about the voltage or current to be supplied to the motor be carried out? The T1 target amplifier (figure 0.1) plays this role by “reading” the computer’s command signals and allowing the commanded amount of voltage or current from the power supply to pass through to the motor.

An amplifier is selected in concert with the power supply and the motor. Common motor types for applications that require position control are brushed DC (BDC) motors, brushless DC (BLDC) motors, and stepper motors. Our T1 target system will have a permanent magnet brushed DC (PMDC) motor, which costs less and requires much simpler electronics than BLDC motors. Stepper motors are also inexpensive and require simple electronics, but they are often inappropriate for continuous and precise motion control and for power-hungry applications.

In chapter 4, motor-driving amplifiers will be considered in detail. For DC motors, they generally fall into two categories:

• Voltage amplifiers, which apply a voltage proportional to a command signal
• Current amplifiers, which apply a current proportional to a command signal

In our T1 target system, we will use a current amplifier.

shows a block diagram of the closed-loop target system. The target computer takes user commands and compares them to encoder measurements. In response to the differences between command and measurement, the controller determines the amount of current that the amplifier should apply to the plant, composed of the motor and flywheel. The lab exercises of each chapter will build up to this closed-loop control, which is fully implemented in lab 7, lab 8.

The motion of the mechanical load of the motor is what we want to control. In particular, we will control the angular velocity and position of a flywheel that is driven by the motor. In lab 4, we will begin controlling the velocity with open-loop control and in lab 7 control the velocity with closed-loop control. Feedback of the angular position is provided by an encoder. Lab 8 explores position control of the flywheel.

Appendix A describes the target electromechanical subsystem in detail.

The prototyping and testing hardware

Additional hardware is used for the prototyping process and testing only. It includes several components that are key to developing the system, debugging issues that arise, and testing. Most components are typical laboratory equipment such as an oscilloscope and a few specialized components such as NAND gate integrated circuits (ICs) for mechanical switch debouncing (see lab 5). Of particular note here is the use of mechanical switches for some prototyping tasks, such as the triggering of finite state machine (FSM) transitions (lab 4) and interrupts (lab 5).

Appendix A describes the prototyping and testing hardware in detail.