SEARCH · Search NASA
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Algorithm for minimizing expected value of quadratic performance index in closed loop optimal control of linear time varying systems.
Recoverability for amplitude and rate bounded optimal control of linear time varying systems
The problem of real time guidance and optimal control of Aeroassisted Orbit Transfer Vehicles (AOTV's) was addressed using singular perturbation theory as an underlying method of analysis. Trajectories were optimized with the objective of minimum energy expenditure in the atmospheric phase of the maneuver. Two major problem areas were addressed: optimal reentry, and synergetic plane change with aeroglide. For the reentry problem, several reduced order models were analyzed with the objective of optimal changes in heading with minimum energy loss. It was demonstrated that a further model order reduction to a single state model is possible through the application of singular perturbation theory. The optimal solution for the reduced problem defines an optimal altitude profile dependent on the current energy level of the vehicle. A separate boundary layer analysis is used to account for altitude and flight path angle dynamics, and to obtain lift and bank angle control solutions. By considering alternative approximations to solve the boundary layer problem, three guidance laws were derived, each having an analytic feedback form. The guidance laws were evaluated using a Maneuvering Reentry Research Vehicle model and all three laws were found to be near optimal. For the problem of synergetic plane change with aeroglide, a difficult terminal boundary layer control problem arises which to date is found to be analytically intractable. Thus a predictive/corrective solution was developed to satisfy the terminal constraints on altitude and flight path angle. A composite guidance solution was obtained by combining the optimal reentry solution with the predictive/corrective guidance method. Numerical comparisons with the corresponding optimal trajectory solutions show that the resulting performance is very close to optimal. An attempt was made to obtain numerically optimized trajectories for the case where heating rate is constrained. A first order state variable inequality constraint was imposed on the full order AOTV point mass equations of motion, using a simple aerodynamic heating rate model.
Control of lightweight flexible arms moving along predefined paths can be successfully synthesized on the basis of a two-time scale approach. A model following control can be designed for the reduced order slow subsystem. The fast subsystem is a linear system in which the slow variables act as parameters. The flexible fast variables which model the deflections of the arm along the trajectory can be sensed through strain gage measurements. For full state feedback design the derivatives of the deflections need to be estimated. The main contribution of this work is the design of an output feedback controller which includes a fixed order dynamic compensator, based on a recent convergent numerical algorithm for calculating LQ optimal gains. The design procedure is tested by means of simulation results for the one link flexible arm prototype in the laboratory.
Optimizing an elastic system and its active control is discussed. Maneuvers from an initial state to a final state in a finite time interval are considered. An active generalized control force that accomplishes the desired maneuver of a prespecified system is optimal if it minimizes a given quadratic cost function. By also varying a set of design parameters, the elastic system can be determined so as to further minimize the cost function. Here, the elastic system that minimizes the actual control cost is compared with the system that minimizes the ratio of actual cost to the cost of optimally maneuvering a rigid system of the same inertial properties. It is shown that an elastic system corresponding to an extremum of the ratio is actually a characteristic of the time-limited maneuver. Because both the spatial domain and the time interval are fixed, a characteristic elastic system is tuned to the specified temporal boundary conditions. The implication for rest-to-rest, spinup, and spin reversal maneuvers of spacecraft is that the optimal control for a characteristic elastic spacecraft is identical to the optimal control for the same spacecraft as if it were rigid.
Not Available
Explore the source record for details and available documents.
Recent advances in optimal control have brought design techniques based on optimization of H(2) and H(infinity) norm criteria, closer to be attractive alternatives to single-loop design methods for linear time-variant systems. Significant steps forward in this technology are the deeper understanding of performance and robustness issues of these design procedures and means to perform design trade-offs. However acceptance of the technology is hindered by the lack of convenient design tools to exercise these powerful multivariable techniques, while still allowing single-loop design formulation. Presented is a unique computer tool for designing arbitrary low-order linear time-invarient controllers than encompasses both performance and robustness issues via the familiar H(2) and H(infinity) norm optimization. Application to disturbance rejection design for a commercial transport is demonstrated.
The goal of this project was to develop and use domain-aware machine learning formulations, based on the Koopman Operator (KO), for modelling multi-scale, multi-disciplinary (e.g., multi-physics), and/or multi-agent systems. The project developed these formulations for the following cases: • Systems with dynamics at two separate time scales, • Systems with a bi-level hierarchical control structure, • Systems with bi-level hierarchical control and dynamics at two separate time scales (the lower level controls operating at the faster time scale), and • Systems with n separate but interacting agents/disciplines (with/without control, respectively); the controls for each agent could include bi-level hierarchical control and dynamics at two separate time scales as described above. The project then defined a set of dynamical systems consisting of different nonlinear oscillators that could be used to test these different formulations and then subsequently learned the KO models for those systems. With the KO models, we were able to do the following: • Quantify system stability, including both long-term and transient behavior, • Quantify the effects of feedbacks between the different time scales and agents/disciplines in terms of those feedbacks’ effects on system stability, • Replace a standard Proportional-Integral (PI) control in the hierarchical control structure with a KO-based Linear-Quadratic Regular (LQR), a form of optimal control, • Calculate optimal supervisory control policies a) with and without time scale separated dynamics at the lower level control levels and b) with both PI and KO-based LQR lower level control policies, and • Calculate dynamic Nash equilibria for multi-agent systems where each agent makes its own control decisions.
The control of linear continuous dynamical systems is investigated as a problem of limited state feedback control. The equations which describe the structure of an observer are developed constrained to time-invarient systems. The optimal control problem is formulated, accounting for the uncertainty in the design parameters. Expressions for bounds on closed loop stability are also developed. The results indicate that very little uncertainty may be tolerated before divergence occurs in the recursive computation algorithms, and the derived stability bound yields extremely conservative estimates of regions of allowable parameter variations.
The reactive collision avoidance (RCA) algorithm allows a spacecraft to find a fuel-optimal trajectory for avoiding an arbitrary number of colliding spacecraft in real time while accounting for acceleration limits. In addition to spacecraft, the technology can be used for vehicles that can accelerate in any direction, such as helicopters and submersibles. In contrast to existing, passive algorithms that simultaneously design trajectories for a cluster of vehicles working to achieve a common goal, RCA is implemented onboard spacecraft only when an imminent collision is detected, and then plans a collision avoidance maneuver for only that host vehicle, thus preventing a collision in an off-nominal situation for which passive algorithms cannot. An example scenario for such a situation might be when a spacecraft in the cluster is approaching another one, but enters safe mode and begins to drift. Functionally, the RCA detects colliding spacecraft, plans an evasion trajectory by solving the Evasion Trajectory Problem (ETP), and then recovers after the collision is avoided. A direct optimization approach was used to develop the algorithm so it can run in real time. In this innovation, a parameterized class of avoidance trajectories is specified, and then the optimal trajectory is found by searching over the parameters. The class of trajectories is selected as bang-off-bang as motivated by optimal control theory. That is, an avoiding spacecraft first applies full acceleration in a constant direction, then coasts, and finally applies full acceleration to stop. The parameter optimization problem can be solved offline and stored as a look-up table of values. Using a look-up table allows the algorithm to run in real time. Given a colliding spacecraft, the properties of the collision geometry serve as indices of the look-up table that gives the optimal trajectory. For multiple colliding spacecraft, the set of trajectories that avoid all spacecraft is rapidly searched on-line. The optimal avoidance trajectory is implemented as a receding-horizon model predictive control law. Therefore, at each time step, the optimal avoidance trajectory is found and the first time step of its acceleration is applied. At the next time step of the control computer, the problem is re-solved and the new first time step is again applied. This continual updating allows the RCA algorithm to adapt to a colliding spacecraft that is making erratic course changes.
Control laws (in some optimal sense) are being developed for the gripper/nut runner end effector. Control laws for the gripper and nut runner portions of the end effector may be developed independently since these two systems are decoupled. A hybrid force/position controller will be used for both the gripper and nut runner. The development of the gripper controller is explained. Sensory data available to the controller is obtained from an array of strain gages as well as a linear potentiometer. Applying well known optimal control theoretic principles, the control which minimizes the transition time between positions is obtained. In addition, a robust force control scheme is developed to contend with the strain gage drift caused by extreme temperature variations encountered in space.
Control-theory design package, Optimal Regulator Algorithms for Control of Linear Systems (ORACLS), developed to aid in design of controllers and optimal filters for systems modeled by linear, time-invariant differential and difference equations. Optimal linear quadratic regulator theory, Linear-Quadratic-Gaussian (LQG) problem, most widely accepted method of determining optimal control policy. Provides for solution to time-in-variant continuous or discrete LQG problems. Attractive to control-system designer providing rigorous tool for dealing with multi-input and multi-output dynamic systems in continuous and discrete form. CDO version written in FORTRAN IV. VAX version written in FORTRAN 77.
Predictor display instrument for optimal manual control, noting use in manned spacecraft missions and time factor
A simple way of tuning usual proportional-and-derivative (PD) controllers is presented. The tuning is devised so that the closed-loop system has a speed of response near that of time-optimal bang-bang control during a large portion of the transient period. When the system is near the desired equilibrium it behaves linearly, and thus avoids the problem of being overly sensitive to the disturbances and noise. The idea is rather intuitive. Unlike dual-mode control, the plant is controlled throughout by a PD controller. Numerical results are included.
In this article, we introduce an adaptive on-line model update algorithm designed for predictive control applications in networked systems, particularly focusing on power distribution systems. Unlike traditional methods that depend on historical data for offline model identification, our approach utilizes real-time data for continuous model updates. This method integrates seamlessly with existing online control and optimization algorithms and provides timely updates in response to real-time changes. This methodology offers significant advantages, including a reduction in the communication network bandwidth requirements by minimizing the data exchanged at each iteration and enabling the model to adapt after disturbances. Furthermore, our algorithm is tailored for non-linear convex models, enhancing its applicability to practical scenarios. The efficacy of the proposed method is validated through a numerical study, demonstrating improved control performance using a synthetic IEEE test case.
This publication is a collection of papers presented at the Mars Mission Research Center workshop on Ongoing Progress in Spacecraft Controls. The technical program addressed additional Mars mission control problems that currently exist in robotic missions in addition to human missions. Topics include control systems design in the presence of large time delays, fuel-optimal propulsive control, and adaptive control to handle a variety of unknown conditions.
A control sensitivity study is conducted for electrified aircraft propulsion concepts using a publicly available turbofan engine simulation. A Proportional-Integral (PI) controller serves as a baseline for comparison against an advanced Model Predictive Control (MPC) approach. MPC is used in the control scheme to allow for subsystem coordination of engine and electric power loads. The electrified turbofan engine includes electric machines to augment the amount of torque being applied to the spools of the engine during transient phases of operation. Power augmentation allows for energy to be stored or exhausted in a manner that increases the fuel efficiency of the engine, reducing aircraft emissions and noise. The study compares the fan speed, high pressure compressor stall margin, and low pressure compressor stall margin against the baseline PI control system, an MPC system without electric machines, and an MPC system with electric machines. The MPC system with electric machines is optimized using the sample time, control horizon and prediction horizon. The MPC design provided a stable damped response as desired and allowed operability margins to be maintained