|The entry of microprocessors into the control loop of control systems has led to flexibility in such systems, and the realization of more advanced control loop algorithms. Though these devices are low cost and offer fast operating frequencies, their computation capability is limited by their basic architecture rather than their frequency. Processors are designed to execute instructions sequentially, which means that the primary way to enhance the control loop response of a processor-based control system is to increase the processor frequency. A type of processor that can perform better than the general purpose processor in a control system is the Digital Signal Processor (DSP). DSPs are specialized processors that work on the same principles as general purpose processors, but offer enhanced arithmetic computation capability. However irrespective of whether a general purpose processor or a DSP is used, there still exists a large array of control applications where both these processors fall far short of real-time requirements.
For such a class of control applications, where the processorfs sequential nature rather than its frequency are causing the real-time bottleneck, we suggest using an FPGA (Field Programmable Gate Array) instead of a fast processor. An FPGA might not be able to operate at the frequency of a fast processor, but it offers the capability of implementing parallelism in design, which can prove to be orders of magnitude faster, and also more reliable than a fast processor solution.
Areas where we have successfully applied FPGAs to real-time control systems include multi-channel sensor data filtering, radar control and data acquisition, motor motion estimation and control, and real-time image filtering etc.
To unleash FPGA on a real-time control problem, we can help get to the solution faster.