Sensitivity analysis of the nonparametric criterion of aircraft flght control system sensors failures detection and isolation

Failures of the aircraft control system sensors can cause both deterioration of stability and controllability characteristics and the inability of safe automatic control. It is necessary to detect and isolate such failures to determine the time and place of their occurrence in order to disable failed sensors or to diagnose them subsequently for reconfiguration during the flight. The direct use of traditional parametric approaches for sensors health monitoring by using their mathematical models is impossible due to the lack of data about the true information input signals received by their sensitive elements. This leads to the necessity of solving the problem of modeling the aircraft flight dynamics with a high level of uncertainties, which makes it difficult to utilize the functional control methods and necessitate the use of excessive sensor hardware redundancy. Well-known nonparametric methods either require a priori knowledge base, preliminary training or long-term tuning on a large volume of real flight data or have low selective sensitivity for reliable detection of failed sensors. In this work, the original nonparametric criterion for detecting and isolating sensors failures is derived. Its sensitivity is analyzed by using a complete nonlinear mathematical model of aircraft flight dynamics with a regular flight control system. The theoretical value and the criterion sensitivity coefficients are determined. The formula for the automatic evaluation of the float criterion threshold value is given. A high convergence of the results with theoretical ones is shown. This makes it possible to use the obtained criterion not only for the instant detection and isolation of sensors failures, but also for preliminary diagnostics of their quantitative characteristics.

The study of metabolic processes in the brain in conditions of hemorrhagic stroke on the background of pharmacotherapy with ademol

In the article, on the model of subarachnoid hemorrhage, some biochemical aspects of the cerebroprotective effect of an industrial sample of an ampoule 1.0% solution of ademol are disclosed, namely its effect on carbohydrate and energy metabolism, the state of antioxidant systems, the activity of lipoperoxidation processes, and the functioning of the L-arginine / NO system in rat brain as possible metabolitotropic components of its protective effect on brain neurons. Severe subarachnoid hemorrhage was created under conditions of propofol anesthesia by injection of heparinized autologous blood (0.1 ml / kg volume) through a catheter located in the subarachnoid space. We used the Student t parametric criterion, W. White nonparametric criterion, Ť Wilcoxon paired criterion — to determine significant changes in the dynamics within the group. Differences were considered statistically significant at p<0.05. Using the example of the acute period of subarachnoid hemorrhage, it was established that ademol is able to eliminate brain energy deficiency (increase the content of adenosine triphosphoric acid and pyruvate in the brain while increasing the energy charge relative to the control pathology samples by an average of 45.1, 42.9 and 22.0%, p<0.05) reduce lactic acidosis (reduce the lactate content by 31.9%, p<0.05), eliminate the manifestations of oxidative stress (reduce the level of malondialdehyde and carbonyl groups of proteins on average by 30.5 and 18.8%, against the background of an increase in the activity of superoxide dismutase, glutathione peroxidase and catalase by 42.1, 25.2 and 37.6%, respectively, p<0.05), simulate the exchange of nitric monoxide (increase the activity of NO synthase with a simultaneous increase in the content of NO L-arginine donor on average by 14.0 and 44.0%, respectively, p<0.05). In these properties, ademol significantly exceeded the effectiveness of solutions of amantadine and magnesium sulfate.

Parameter Estimation of Lotka-Volterra Model : A Two-Step Model

The deterministic power law logistic model is used to describe density-dependent population growth for cases when ordinary logistic model is found to be insufficient. This paper estimates the parameters of stochastic power law logistic model specifically the Lotka-Volterra model by employing the two-step approach. The Bayesian approach is implemented in the first step of estimating the regression spline parameters. Combining the existing and proposed nonparametric criterion, the structural parameters of SDE are estimated in the second step. Results indicate high percentage of accuracy of the estimated diffusion parameter of Lotka-Volterra model supporting the adequacy of the proposed criterion as an alternative to the classical methods.