The Aim of this paper consists in the development of a method of improving the functional dependability of the control systems of unmanned aerial vehicles (UAV CS) affected by electromagnetic effects in flight and failures within the functional component of the onboard test instrumentation (OBTI).That is achieved through the identification of the failed functional element, the functional component of OBTI, the capability of performing the target objective of the UAV CS and decision-making regarding the initiation of the flexible operation algorithm. The existing and future UAV CS under development use binary reliability models, i.e. two states are distinguished: up and disabled. Therefore, each in-flight failure is classified as the UAV CS failure regardless of the current mission. If we regard a CS as a multifunctional system, it becomes obvious that the failure of not any UAV CS functional element causes flight termination. Methods. Solving the problem involved the use of a CS diagnostic model in the form of binary relations between the control actions and combinatorial subsets of functional elements, risk of losses estimation method as part of improving the functional dependability of UAV CS in flight, decision theory and combined branch-and-bound method. The mission performance probability is used as the efficiency criterion. This criterion is applicable when changes in a UAV CS’ characteristics cause only partial reduction of the functional efficiency. Results. The purpose of CBTI self-supervision is failure location with the depth that allows determining its ability to perform the basic operations with the probability not lower than required by the customer, as well as the allowed set of elementary checks (EC) in this case. Based on the current results of elementary self-checks (ESC), one of the following decisions can be taken: stop the checks and discard CBTI; continue location; stop failure location and continue UAV CS mission per modified algorithm. At each stage of failure location in CBTI, based on the results of ESC, the area of covering check (ACC) and part of set suspected of failure (PSSF) are analyzed, which includes verifying the ACC for sufficient coverage of the PSSF, based on which appropriate decisions are taken. The following areas are formed: the area of observable data (processes of changes in the ACC and PSSF areas), within which the decision is taken to continue the checks, and the area, within which it is finally decided to terminate the checks. If It Is decided to continue the failure location, another ESC is selected, which Is associated with the risk of loss. The probability of false discarding of CBTI due to ESC selected out of ACC Is taken as the risk of loss. The moment of termination of CBTI self-supervision depends not only on the set of decisions, but their sequence as well. Thus, the task at hand comes down to designing the optimal ESC strategy that minimizes the probability of false discarding. The idea of combined branch-and-bound method (CBBM) as part of the design of the optimal CBTI self-supervision algorithm consists in the consecutive selection at each stage of ESC implementation process, out of the subset of minimum risk checks of the next ESC till a one-element subset is obtained and/or the corresponding decision is taken. Conclusions. The developed method allows continuing the performance of the target objectives of a UAV CS In flight when affected by failures In CBTI.
Application of multi-agent approach in the electric power control systems within the active energy complex
