THE ALGORITHM FOR ESTIMATING ANGULAR RATE OF SPACECRAFT IN SURVIVABILITY MODE

Texto integral

Introduction. Spacecraft (SC) operation in orbit is connected to such an important concept as permanent performing of an object function. In the early going, the quality management system of SC is meant to provide SC runnability in case of equipment failure [1; 2]. For their part, the failures can be classified as: a) failures not hav- ing a considerable effect on performing an object function of SC and b) failures capable of leading to a temporary time gap in the operation of SC for the purpose intended [3]. The time gap can appear due to loss of orientation. For SC, engaged in communication session, in the areas of lack of communication, disorientation and electric battery discharge can cause disastrous effects, e.g. loss of spacecraft. Survivability mode. To provide survivability of space- craft during its continuous autonomous operation in conoccurred on the geostationary SC engineered by JSC “ISS - Reshetnev”. Complete failure of an angular rate sensor would lead to disability of using survivability mode. An appropriate solution can be the algorithm of calculating angular rate on available orienting point - a star or the Sun, which is offered in [8]. However, using of these ways does not give the opportunity to provide search speed without orienting point in sight of a solar sensor. We offered the solution in the form of a control algo- rithm taking into account failures of an angular rate sensor via one or more control channels with additional protection from short time false indications. To define angular rate, we introduced an additional block, which calculates evaluation of angular rate accord- ing to the following formula: ditions of outer space, JSC “ISS - Reshetnev” use autono- Wˆ i = Wop + Wor.eng , (3) mous control loop of vehicle orientation - survivability i -1 i -1 W mode [4-6]. According to a patent, the idea of the mode where Wˆ i is the calculating values of angular rate at preinvolves automatic unfolding of a solar panel into a fixed position relative to a main body of spacecraft to make sent, ′/s; op i -1 - control rate at a previous step, ′/s; solar panels as enlightened by the Sun as possible. Conor.eng W i -1 - calculated value change of angular rate of SC sequently, spacecraft maintains positive power balance and it is saved. The flow chart of an autonomous control loop in survivability mode is shown in the fig. 1. according to generated control action speed at a previous step, ′/s. In survivability mode, ignition of an orientation engine provides control input according to the formula: The operation algorithm consists of three blocks: Wor.eng = t × p × l / J , (4) 1) defining the angle of orientation of SC (Ψ = i -1 i -1 = [ψ φ θ], ′), which is performed using a solar sensor (SS) [7]. In that mode, we used solar sensors that identify the sun position by means of two angles (φ and θ); 2) defining the angular rate of SC (Ω = [ωx ωy ωz], ′/s) it is performed using an angular-rate sensor or using the algorithm of angular rate calculation [8]; 3) calculation and generation of control actions on SC [9]. The unit of evaluation and generation of control actions on SC exerts control by the laws [10]: é-Kay - Kdw ù ê x x x ú where τi-1 is the time of orientation engine ignition, s; p - rated thrust of an orientation engine, Н; l - the arm of an orientation engine action, m; J - momentum of inertia of SC, kg·m2. Using of computer model of orientation engine ignition allows calculating the angular rate of SC that provides the system with parameters for controllability. The inaccu- racy of angular rate depends on the accuracy of computer model of an orientation engine, as well as on initial value of angular rate. Nevertheless, using of estimation of angu- lar rate by generated action allows providing search rotay y y M = -KaY - K DW = ê-Kaj - Kdw ú , z z z û êëê -Kaq - Kdw úú (1) tion of SC in the case of failure of all angular rate sensors that measure angular rate of SC in direction of search rotation. x ˆ op ˆ meas ˆ where Ka = [ Ka τ = J / M · Δt, (2) K z y a Ka ] is the coefficient of amplifica- Using a filter of angular rate of SC helps solve the task of protection from false indicated readings of rateme- ter (levelling of impact) and compensation of inaccuracy tion, N·m·s/′; Kd = [ Kd Kd Kd ] - damping coefficient, of defining angular rate by the impact: Подпись: x Подпись: y Подпись: z N·m·s2/′; М - required control moment, N·m; τ - total Wi = Wi + k (Wi - Wi ), (5) W time of generation of control signal, s; Δt - control period, s; J - momentum of inertia of SС, kg·m2. where Wˆ op is control angular rate at present step, ′/s; i Using of orientation sensors (solar sensors and angular meas i - measured angular rate, ′/s; k - filter gain. The rate sensors), on-board digital computer system and jet orientation engines supports the survivability mode. Par- tial or complete failure of an angular rate sensor can cause a problem. Partial failure of an angular rate sensor filter gain defines the convergence rate of rate estimation to measured rate, and it is chosen in the range from 0 to 1. In fig. 2 we presented The flow chart of improved autonomous control loop in survivability mode. Fig. 1. Flow chart of an autonomous control loop in survivability mode Рис. 1. Функциональная схема автономного контура управления в режиме живучести Fig. 2. Flow chart of improved autonomous control loop in survivability mode Рис. 2. Функциональная схема усовершенствованного автономного контура управления в режиме живучести Tests. For algorithm qualification, we used mathe- matical modeling in mathematical computation environment Octave GNU and in simulation ground-based debugging complex of onboard radioelectronics [11]. The mathe- matical model consisted of the following blocks: kinematic and dynamic model of SC [12]; device model; con- trol system model [13]; model of orientation engine [14]. We simulated situations when at starting moment of time, angular rate in onboard software and model angular rate differed significantly (from -15 to 15 ′/s) for different failures of angular rate sensors (8 failure combinations). We carried out more than 1000 test variations. Fig. 3 shows the results of two tests for two cases: a - failure of an angular rate sensor via channel X, b - failure of an angular rate sensor via all available channels. In the fig. 3 we labelled model angular rate of SC - ωxmod and angu- lar rate estimation - ωx est. Readings comparison of model angular rate and angu- lar rate estimation shows ability to make required control actions via all control channels. We should focus on the situation of complete failure of an angular rate sensor and, consequently, lack of information about angular rate of SC. In this variant of failure, the algorithm of angular rate estimation solves the task of defining angular rate. Angu- lar rate decrease via roll channel is impossible in some modes; one of these situations is shown in fig. 3, b. Its analysis shows that angular rate via pitch channels and yaw channels are calculated with some inaccuracy that decreases over time. The initial rate via roll channel (un- known to the algorithm) is not decelerated but it does not get lower. The fact that orientation is performed by a roll axis to the Sun does not give a possibility to measure angular rate via a roll channel using this orienting point. As high reliability and maximum possible development of the algorithm are required, we performed additional checking on an “iron bird”, which consisted of real tech- nological devices of a solar sensor and an angular rate sensor installed on rotary tables. You can see the descrip- tion of an “iron bird” and methods of testing in the sources [14; 15]. We presented the graphs of angular rate changing while “iron bird” testing in fig. 4. We can describe the results of the ground processing as effective: the algorithm successfully passed all the tests in all variants for all possible combinations of failures of an angular-rate sensor at different stages of processing. The final test of the algorithm was live testing carried out on real SC, i. e. we prepared the programme to take spacecraft off into survivability mode with the simulation of angular-rate sensor failures. We carried out the four tests successfully. The logic of survivability mode with the improved algorithm effectively dealt with the orienta- tion of SC on the Sun in the context of simulated failures of an angular rate sensor. Graphs of angular rate changing at verification tests in situ are shown in fig. 5. а b Fig. 3. Results of ground-based modeling tests with (a) failure of an angular rate sensor via channel X; failure of an angular rate sensor via all available channels (b) Рис. 3. Результаты испытаний при наземном моделировании при отказе только ДУС по каналу X (а), и отказе ДУС по всем каналам управления (б) Fig. 4. Graphs of angular rate changing while “iron bird” testing Рис. 4. Графики изменения угловой скорости при полунатурных испытаниях Fig. 5. Graphs of angular rate changing at verification tests in situ Рис. 5. Графики изменения угловой скорости при натурных испытаниях Conclusion. As a result of the research performed, the introduced method of algorithm improvement allows counteracting eventual failures of angular-rate sensors and ensuring protection from short time false indications; it was proved by the results of ground processing, “iron bird” tests and flight tests. The improved algorithm is effectively applied on the geostationary SC developed by the JSC ISS - Reshetnev. However, the algorithm has some drawbacks. One of them is impossibility to define a full vector of angular rate. Another variant of failures may be the failure of a solar sensor. We did not consider it in the paper, but it is rather important. In this case, the appropriate solution to the problem can be orientation in the current from solar panels.

Sobre autores

S. Latyntsev

JSC “Academician M. F. Reshetnev” Information Satellite Systems”

Email: lat.sv@mail.ru
52, Lenina Str., Zheleznogorsk, Krasnoyarsk region, 662972, Russian Federation

S. Meus

JSC “Academician M. F. Reshetnev” Information Satellite Systems”

52, Lenina Str., Zheleznogorsk, Krasnoyarsk region, 662972, Russian Federation

А. Оvсhinniкоv

JSC “Academician M. F. Reshetnev” Information Satellite Systems”

52, Lenina Str., Zheleznogorsk, Krasnoyarsk region, 662972, Russian Federation

Bibliografia

Arquivos suplementares