Intelligent power management system with fault detection capabilities for armored fighting vehicles

Electric machines are widely used in almost all aspects of our lives. Due to starting of heavy machines or a group of small machines simultaneously, the electric systems face the problem of voltage fluctuations. In addition to this, when a number of machines are turned ON together, it may so happen that the overall power requirements exceed the power source capacity. Consequently, it is not possible to switch ON all machines at the same time. Such scenarios become very critical in systems like an armored fighting vehicle where the machines or loads to be switched ON are crucial and the power supply is limited by the amount of fuel carried in the vehicle. Under these conditions, there is a need for a system that can automatically manage the distribution of source power among various loads so as to optimize the fuel utilization and performance of the vehicle. Most of the existing methods targeted to solve these problems, use different types of converters, which makes the circuit complex. Another problem with the electric systems is occurrence of faults in electric machines which may be because of many electrical or mechanical imbalances. Fault detection is the first step toward avoiding the failure of the system due to the occurrence of various faults. The problem of fault detection in electric machines under stationary operating conditions is addressed in various works, but little has been reported about fault detection under non-stationary conditions. This work presents an Intelligent Power Management System (IPMS) which provides a solution to the problem of power management and voltage fluctuations without increasing system complexity and a fault detection framework for detecting faults in electric dc machines. The proposed IPMS is designed to address these problems in AFVs and can be generalized for any application scenario. For fault detection in conventional dc machines, this work presents a framework to identify the unique signatures corresponding to the various faults not only under stationary, but under non-stationary operating conditions as well. From these signatures, a model can be trained to predict any abnormal condition before its occurrence. Finally, the results obtained by simulating these problems and implementing proposed IPMS and fault detection framework present their efficacy.