This paper describes a new fault-tolerant control technique for over-actuated uncertain linear systems that compensates for actuator faults and failures using an output integral sliding mode-based control allocation strategy. An observer-based fault estimation unit is first proposed to estimate the system states and the actuator’s effectiveness level. Based on the estimated state’s information, a nominal virtual control law is created to achieve the desired specification of the perturbed system. A nonlinear output integral sliding manifold is incorporated with the nominal virtual control law that provides resilience to the closed-loop system against the uncertainty caused by the actuator’s faults and failure, states estimation error, and fault estimation error. Finally, depending on anticipated actuator efficacy from the fault estimation unit, the control allocation reroutes the virtual control input signals among the redundant actuators. A small-gain theorem is used to demonstrate the augmented closed-loop system stability. The observer and controller gains are synthesized using the linear matrix inequality technique. Finally, simulations on an aircraft system are performed to verify the efficacy of the suggested FTC technique. Compared with the existing work, the proposed approach is better able to handle actuator redundancy in faulty conditions and cater to the fault estimation error up to a certain level. Note to Practitioners —The aim of this paper is to address the reliability issues of industrial systems that contain sufficient input redundancy and are particularly designed to tolerate the faults and failures issues during the system operation. Existing fault-tolerant control schemes based on control allocations have limitations in terms of effective utilization of actuator redundancy, handling of fault estimation error, and applicability of reconfigurable control law. In this paper, we developed a fault-tolerant control strategy that is capable of addressing all the aforementioned concerns. The output integral sliding mode control-based control allocation scheme, proposed in this paper, is applicable to the class of over-actuated systems. The reconfiguration of fault-tolerant control law based on the estimated states and fault information is capable to cover a wider class of faults and failures while maintaining robustness against the system dynamics and uncertainty. The control reconfiguration is equally applicable to the class of systems that contains both rank-deficient and in-deficient input distribution matrices. The proposed scheme assumed the bounded external disturbance, therefore in future work, the adaptive law will be incorporated with sliding mode controller and results will be extended to a generalized class of affine nonlinear systems.

Heating, Ventilation and Air Conditioning (HVAC) systems account for 59% of energy consumption in domestic buildings and 36% in non-domestic buildings. According to a study, around 39% of occupants are dissatisfied with indoor temperature in non-domestic buildings. To maintain thermal comfort and indoor air quality, HVAC systems are widely used in non-domestic buildings. This research aims to develop energy-efficient control techniques for HVAC systems while ensuring indoor thermal comfort. Three control strategies, namely EnergyPlus model-based Model Predictive Control (MPC), Sliding Mode Control (SMC), and simple ON/OFF control, are employed and compared at the Department of Electrical and Computer Engineering, COMSATS University Islamabad, Lahore Campus. Furthermore, a machine learning-based Predicted Mean Vote (PMV)-based temperature setpoint estimator is designed to ensure occupant thermal comfort. The control techniques estimate the temperature setpoints and supply air temperature of the Variable Air Volume (VAV) system to control indoor room temperature. The energy consumption and indoor thermal comfort of the building are compared under different control techniques. The results show that MPC with PMV-based setpoints consumes 17.20% less energy during winters and 14.67% less energy during summers than a simple ON/OFF controller.

An integrated integral sliding mode fault tolerant control scheme is proposed for the linear parameter varying (LPV) plants in the output feedback framework. To estimate the unavailable plant states, an unknown input LPV observer is employed. The feedback and observer gains are designed using linear matrix inequalities approach to ensure closed‐loop stability nominally and in the situation of actuator faults/failures. An integral sliding modes approach in the LPV framework together with control allocation maintains the nominal performance in the presence of permissible set of actuator faults/failures. A rigorous closed‐loop stability analysis is carried out and to show the feasibility of the proposed approach it is tested in simulations on the non‐linear plant of a passenger aircraft.

This paper proposes a novel integral sliding mode on-line control allocation scheme for fault tolerant control. The order of the sliding motion in the sliding mode is not reduced as in conventional sliding mode schemes. Also the robustness of the closed loop system can be guaranteed throughout the entire response, starting from the initial time instant, because there is no reaching phase. Actuator effectiveness levels are used to redistribute the control signals to healthy actuators online in the case of faults or failures without reconfiguring the controller. In this way the scheme can cope with total failure of certain actuators. Simulation results using the Admire aircraft model show good performance in fault and even certain total actuator failure scenarios which validates the effectiveness and feasibility of the proposed scheme.