A high proportional gain results in a large change in the output for a given change in the error. If the proportional gain is too high, the system can become unstable (see the section on loop tuning). In contrast, a small gain results in a small output response to a large input error, and a less responsive or less sensitive controller. If the proportional gain is too low, the control action may be too small when responding to system disturbances. Tuning theory and industrial practice indicate that the proportional term should contribute the bulk of the output change.[citation needed]
Handbook Of Pi And Pid Controller Tuning Rules Pdf 13
Download File: https://tinourl.com/2vKOe5
The integral term accelerates the movement of the process towards setpoint and eliminates the residual steady-state error that occurs with a pure proportional controller. However, since the integral term responds to accumulated errors from the past, it can cause the present value to overshoot the setpoint value (see the section on loop tuning).
Designing and tuning a PID controller appears to be conceptually intuitive, but can be hard in practice, if multiple (and often conflicting) objectives, such as short transient and high stability, are to be achieved. PID controllers often provide acceptable control using default tunings, but performance can generally be improved by careful tuning, and performance may be unacceptable with poor tuning. Usually, initial designs need to be adjusted repeatedly through computer simulations until the closed-loop system performs or compromises as desired.
Mathematical PID loop tuning induces an impulse in the system and then uses the controlled system's frequency response to design the PID loop values. In loops with response times of several minutes, mathematical loop tuning is recommended, because trial and error can take days just to find a stable set of loop values. Optimal values are harder to find. Some digital loop controllers offer a self-tuning feature in which very small setpoint changes are sent to the process, allowing the controller itself to calculate optimal tuning values.
While PID controllers are applicable to many control problems, and often perform satisfactorily without any improvements or only coarse tuning, they can perform poorly in some applications and do not in general provide optimal control. The fundamental difficulty with PID control is that it is a feedback control system, with constant parameters, and no direct knowledge of the process, and thus overall performance is reactive and a compromise. While PID control is the best controller in an observer without a model of the process, better performance can be obtained by overtly modeling the actor of the process without resorting to an observer.
The form of the PID controller most often encountered in industry, and the one most relevant to tuning algorithms is the standard form. In this form the K p \displaystyle K_p gain is applied to the I o u t \displaystyle I_\mathrm out , and D o u t \displaystyle D_\mathrm out terms, yielding:
Most commercial control systems offer the option of also basing the proportional action solely on the process variable. This means that only the integral action responds to changes in the setpoint. The modification to the algorithm does not affect the way the controller responds to process disturbances.Basing proportional action on PV eliminates the instant and possibly very large change in output caused by a sudden change to the setpoint. Depending on the process and tuning this may be beneficial to the response to a setpoint step.
When these controllers are tuned appropriately, they can ensure sufficient performance in most of the required UAV applications, including demanding active interaction with the environment. Furthermore, they neither require the knowledge of the analytical model of UAVs nor they require performing any online estimation of the parameters. The introduction of an appropriate optimality criterion in the form of a cost function streamlines the tuning procedure and synthesis of a controller. In the problem considered in this paper, it is assumed, in addition, that the tuning procedure of a controller (or performing fine-tuning for rough initial estimates of controller parameters) is performed in real time during flight of a UAV. This is especially useful in applications such as transporting load by a single flying unit [7,8] or by multiple cooperating UAVs [9] that require even better precision of trajectory tracking due to the coupling between the robots, e.g., in grasping tasks (see Figure 1). When the load changes, it is important to tune controller parameters to the new dynamics of the system. In the presented approach, it is also possible in the ideal case of no model-UAV mismatch (perfect modeling), to obtain an optimal controller parameters for various flight scenarios, e.g., in acrobatic maneuvres (with dynamic acceleration phase) or maximization of flight time by avoiding sudden moves using smooth and feasible reference trajectories.
The necessity of frequent controller tuning has been highlighted in the MBZIRC competition, which required a fleet of multiple UAVs with different sensory equipment and actuators, but its importance seems as big in any deployment of UAV systems. Very precise manual tuning of one set of controller parameters takes days of flying outdoor in different environment conditions. Moreover, each change of the platform (mainly those that influence UAV weight) requires obtaining a new set of parameters and therefore additional time-consuming on flight parameter analyses is needed. Therefore, the aim of this paper is to design an advanced method for parameter tuning that achieves even better results than manual tuning in a fraction of time required by current approaches.
In this paper, the authors propose an alternative approach to the methods listed in the above paragraph to perform automated tuning of controllers, using deterministic optimization methods. It is based on the assumption that the obtained solution for a specific type of a cost function is a global minimizer with prescribed tolerance, calculated in a defined time regime. From multiple available approaches to tuning problems, easily-implementable iterative optimization methods have been selected and expected to allow either tuning or fine-tuning of controllers during flight of UAVs. Their computational complexity and conceptual simplicity is an advantage in comparison with multi-agent based optimization techniques, where time necessary to conduct all computations is proportional to the number of agents. The approach in this paper is inspired by a classic Iterative Feedback Tuning method [23,24,25] in which optimal controller parameters are sought in reference to some cost function calculated using a current output signal of a closed-loop system. The references mention the latter method among such approaches as Fictitious Reference Iterative Tuning [26,27,28], Iterative Learning Approach [29,30] or Virtual Reference Feedback Tuning [31]. The latter methods enable direct tuning based on experimental input-output data. Direct methods can be divided into classes such as zero- or first-order optimization techniques. The first class uses only zero-order interval arithmetic, as in the bisection, golden-search or Fibonacci-search method, and the second one uses interval gradients or interval slopes, as in IFT or ILA, Steepest Descent Method, Reinforcement Learning [32] or methods proposed in [30,33].
Ref. [39] presents a novel strategy of maximum power point tracking for photovoltaic power generation systems based on the Fibonacci-search technique. The aim of the paper has been to realize a simple control system to track the real maximum power point even under non-uniform or for rapidly changing insolation conditions. In the article [40], a practical PI/PID controller tuning method for integrating processes with dead time and inverse response based on a model has been proposed. Optimum tuning of the parameter for disturbance rejection based on the model and the minimum of the IAE criterion has been successfully obtained via adaptation of the golden-search technique. However, the approach presented in the current paper aims to result in obtaining an iterative tuning scheme of controller parameters independently of the availability of the model, as shown in [41], where, to the best knowledge of the authors, there is only a single reference available in the literature referring to the concept of using Fibonacci-search method in iterative model-based PID controller tuning. Ref. [41] unfortunately delivers only a blurry and general procedure for possible search in parameter space stating that the reciprocated Fibonacci sequence which has been used to adjust the controller parameters has been chosen arbitrarily because it is a convenient, convergent sequence. The paper does not present in full the potential and possible advantages of the use of this algorithm in controller tuning procedures, which is the basic reason for writing this paper and formulating further research.
Recalling research motivation from Section 3, the proposed method to perform auto-tuning would be even more suitable in the case of fine-tuning of nominal controller parameters aiming at improving its responsive capabilities to different dynamics requirements than to its prototyping from a scratch. The latter is the results of no need to have an analytical criterion using knowledge from a model of UAV, allowing for obtaining a good estimate of the search space for the given controller, ensuring stability of the control system. From the implementation point of view, and in order to ensure the safety of the tuning procedure, it is recommended to start the tuning from narrow deviations with respect to the nominal controller parameters and widen then along the way, repeating the procedure, if necessary.
Sample snapshots of one of the experiments conducted in ROS: (a,b) UAV takes off from the ground and flies to the desired position in an autonomous mode; (c) safe altitude is reached (in desired coordinates); (d) decreasing the height down to 2m; (e,f) altitude controller (PD) autotuning: repeated 48 cycles of raising up to 3m and going down again the level of 2m (12s per single iteration); (g) termination of the autotuning experiment with autonomous landing of the UAV on the ground ( ). 2ff7e9595c
Comments