Abstract
This paper addresses the synthesis and optimisation of an adaptive recursive digital controller for the roll-angle control loop of a quadcopter based on a Sugeno-0 fuzzy supervisor. The relevance of the study is determined by the growing need to improve the dynamic performance and energy efficiency of quadcopter flight-control systems equipped with brushless DC motors. In practical flight conditions, the external roll loop must provide fast and accurate response under variable manoeuvring regimes while remaining sufficiently simple for real-time embedded implementation. However, conventional controllers with fixed coefficients usually ensure acceptable control quality only within a limited operating range. An increase in response speed is commonly accompanied by larger overshoot, higher peak currents, and greater energy consumption. Therefore, the problem considered in the paper is the development of a control law capable of adapting its action intensity according to the current tracking conditions without excessive computational complexity. The aim of the study is to develop, mathematically substantiate, and numerically investigate an adaptive recursive digital controller for the roll-angle loop of a quadcopter in which the Sugeno-0 supervisor directly scales the controller parameters of the FIR numerator, and to compare this approach with previously implemented control solutions in terms of transient quality and energy consumption. The proposed controller combines a fixed IIR denominator obtained by deadbeat synthesis with an adaptive FIR numerator whose coefficients are updated online at each sampling instant by fuzzy inference. In contrast to approaches based on intermediate parameter adaptation followed by online controller resynthesis, the proposed structure directly modifies the recursive controller coefficients, which simplifies practical implementation. The research methods include the development of a physically grounded mathematical model of the BLDC motor and the one-axis roll platform of the quadcopter, continuous-to-discrete transformation of the plant equations, deadbeat synthesis of the denominator, pole assignment using the Ackermann approach, fuzzy supervision based on the normalized roll-angle error and its increment, and stability assessment by the frozen-coefficients method and the Jury criterion. The paper also provides physical justification of the Sugeno-0 rule tables and visualises the supervisor behaviour by means of output surfaces. The results of comparative simulation for four controllers in nine test scenarios show that the proposed adaptive recursive digital controller provides an improved compromise between response speed, overshoot limitation, and energy expenditure. The obtained results confirm that direct fuzzy adaptation of the FIR numerator coefficients makes it possible to improve roll-angle control quality while preserving computational simplicity. It is concluded that the proposed controller is a promising solution for embedded quadcopter flight-control systems operating under variable conditions.