Dynamic diagnostic and reconfiguration for a new FPGA based green controller

In order to apply the CPU to the control system for the renewable energy plant which is usually installed in the isolated place, there are a lot of issues e.g. the cooling, the power conservation and the parts exchange with the device life cycle. The undetectable dangerous error of the control system is also a big issue for such the isolated plant with the long term operation, because it causes the decline of the effectiveness of the redundancy for the reliability and the availability. In order to address the undetectable error, it is effective to enhance the diagnostic with such an approach as the internal process is compared as well as the output. It is not a realistic solution that the control system based on the CPU conducts such diagnostic, because such the heavy diagnostic requires the considerable high performance of the CPU. Its exhaust heat may cause the necessity of the additional cooling facility. The FPGA based green controller is able to calculate such the heavy diagnostic also in parallel. It does not influence to the calculation performance of the control. This paper presents a new architecture of the FPGA based green controller of which the reliability is enhanced by the parallel diversity diagnostic. Furthermore, it presents a new approach of the realization of the diversity for FGPA. It is confirmed by the FPGA simulation that the proposed diversity approach improves the error detection efficiency.