Architecture
Self-tuning systems are typically composed of four components: expectations, measurement, analysis, and actions. The expectations describe how the system should behave given exogenous conditions.
Measurements gather data about the conditions and behaviour. Analysis helps determine whether the expectations are being met- and which subsequent actions should be performed. Common actions are gathering more data and performing dynamic reconfiguration of the system.
Self-tuning (self-adapting) systems of automatic control are systems whereby adaptation to randomly changing conditions is performed by means of automatically changing parameters or via automatically determining their optimum configuration. [2] In any non-self-tuning automatic control system there are parameters which have an influence on system stability and control quality and which can be tuned. If these parameters remain constant whilst operating conditions (such as input signals or different characteristics of controlled objects) are substantially varying, control can degrade or even become unstable. Manual tuning is often cumbersome and sometimes impossible. In such cases, not only is using self-tuning systems technically and economically worthwhile, but it could be the only means of robust control. Self-tuning systems can be with or without parameter determination.
In systems with parameter determination the required level of control quality is achieved by automatically searching for an optimum (in some sense) set of parameter values. Control quality is described by a generalised characteristic which is usually a complex and not completely known or stable function of the primary parameters. This characteristic is either measured directly or computed based on the primary parameter values. The parameters are then tentatively varied. An analysis of the control quality characteristic oscillations caused by the varying of the parameters makes it possible to figure out if the parameters have optimum values, i.e.. if those values deliver extreme (minimum or maximum) values of the control quality characteristic. If the characteristic values deviate from an extremum, the parameters need to be varied until optimum values are found. Self-tuning systems with parameter determination can reliably operate in environments characterised by wide variations of exogenous conditions.
In practice systems with parameter determination require considerable time to find an optimum tuning, i.e. time necessary for self-tuning in such systems is bounded from below. Self-tuning systems without parameter determination do not have this disadvantage. In such systems, some characteristic of control quality is used (e.g., the first time derivative of a controlled parameter). Automatic tuning makes sure that this characteristic is kept within given bounds. Different self-tuning systems without parameter determination exist that are based on controlling transitional processes, frequency characteristics, etc. All of those are examples of closed-circuit self-tuning systems, whereby parameters are automatically corrected every time the quality characteristic value falls outside the allowable bounds. In contrast, open-circuit self-tuning systems are systems with para-metrical compensation, whereby input signal itself is controlled and system parameters are changed according to a specified procedure. This type of self-tuning can be close to instantaneous. However, in order to realise such self-tuning one needs to control the environment in which the system operates and a good enough understanding of how the environment influences the controlled system is required.
In practice self-tuning is done through the use of specialised hardware or adaptive software algorithms. Giving software the ability to self-tune (adapt):
- Facilitates controlling critical processes of systems;
- Approaches optimum operation regimes;
- Facilitates design unification of control systems;
- Shortens the lead times of system testing and tuning;
- Lowers the criticality of technological requirements on control systems by making the systems more robust;
- Saves personnel time for system tuning.
Control theory is a field of control engineering and applied mathematics that deals with the control of dynamical systems in engineered processes and machines. The objective is to develop a model or algorithm governing the application of system inputs to drive the system to a desired state, while minimizing any delay, overshoot, or steady-state error and ensuring a level of control stability; often with the aim to achieve a degree of optimality.
In computer science, a search algorithm is an algorithm designed to solve a search problem. Search algorithms work to retrieve information stored within particular data structure, or calculated in the search space of a problem domain, with either discrete or continuous values.
In computer science and operations research, a genetic algorithm (GA) is a metaheuristic inspired by the process of natural selection that belongs to the larger class of evolutionary algorithms (EA). Genetic algorithms are commonly used to generate high-quality solutions to optimization and search problems by relying on biologically inspired operators such as mutation, crossover and selection. Some examples of GA applications include optimizing decision trees for better performance, solving sudoku puzzles, hyperparameter optimization, causal inference, etc.
Mathematical optimization or mathematical programming is the selection of a best element, with regard to some criterion, from some set of available alternatives. It is generally divided into two subfields: discrete optimization and continuous optimization. Optimization problems arise in all quantitative disciplines from computer science and engineering to operations research and economics, and the development of solution methods has been of interest in mathematics for centuries.
Reinforcement learning (RL) is an interdisciplinary area of machine learning and optimal control concerned with how an intelligent agent ought to take actions in a dynamic environment in order to maximize the cumulative reward. Reinforcement learning is one of three basic machine learning paradigms, alongside supervised learning and unsupervised learning.
In computational intelligence (CI), an evolutionary algorithm (EA) is a subset of evolutionary computation, a generic population-based metaheuristic optimization algorithm. An EA uses mechanisms inspired by biological evolution, such as reproduction, mutation, recombination, and selection. Candidate solutions to the optimization problem play the role of individuals in a population, and the fitness function determines the quality of the solutions. Evolution of the population then takes place after the repeated application of the above operators.
Performance tuning is the improvement of system performance. Typically in computer systems, the motivation for such activity is called a performance problem, which can be either real or anticipated. Most systems will respond to increased load with some degree of decreasing performance. A system's ability to accept higher load is called scalability, and modifying a system to handle a higher load is synonymous to performance tuning.
A control system manages, commands, directs, or regulates the behavior of other devices or systems using control loops. It can range from a single home heating controller using a thermostat controlling a domestic boiler to large industrial control systems which are used for controlling processes or machines. The control systems are designed via control engineering process.
In computational science, particle swarm optimization (PSO) is a computational method that optimizes a problem by iteratively trying to improve a candidate solution with regard to a given measure of quality. It solves a problem by having a population of candidate solutions, here dubbed particles, and moving these particles around in the search-space according to simple mathematical formula over the particle's position and velocity. Each particle's movement is influenced by its local best known position, but is also guided toward the best known positions in the search-space, which are updated as better positions are found by other particles. This is expected to move the swarm toward the best solutions.
In computer science and mathematical optimization, a metaheuristic is a higher-level procedure or heuristic designed to find, generate, tune, or select a heuristic that may provide a sufficiently good solution to an optimization problem or a machine learning problem, especially with incomplete or imperfect information or limited computation capacity. Metaheuristics sample a subset of solutions which is otherwise too large to be completely enumerated or otherwise explored. Metaheuristics may make relatively few assumptions about the optimization problem being solved and so may be usable for a variety of problems.
In the design of experiments, optimal experimental designs are a class of experimental designs that are optimal with respect to some statistical criterion. The creation of this field of statistics has been credited to Danish statistician Kirstine Smith.
In computer science and operations research, Genetic fuzzy systems are fuzzy systems constructed by using genetic algorithms or genetic programming, which mimic the process of natural evolution, to identify its structure and parameter.
Adaptive control is the control method used by a controller which must adapt to a controlled system with parameters which vary, or are initially uncertain. For example, as an aircraft flies, its mass will slowly decrease as a result of fuel consumption; a control law is needed that adapts itself to such changing conditions. Adaptive control is different from robust control in that it does not need a priori information about the bounds on these uncertain or time-varying parameters; robust control guarantees that if the changes are within given bounds the control law need not be changed, while adaptive control is concerned with control law changing itself.
In control theory, advanced process control (APC) refers to a broad range of techniques and technologies implemented within industrial process control systems. Advanced process controls are usually deployed optionally and in addition to basic process controls. Basic process controls are designed and built with the process itself, to facilitate basic operation, control and automation requirements. Advanced process controls are typically added subsequently, often over the course of many years, to address particular performance or economic improvement opportunities in the process.
IOSO is a multiobjective, multidimensional nonlinear optimization technology.
In mathematical optimization, the firefly algorithm is a metaheuristic proposed by Xin-She Yang and inspired by the flashing behavior of fireflies.
OptiY is a design environment software that provides modern optimization strategies and state of the art probabilistic algorithms for uncertainty, reliability, robustness, sensitivity analysis, data-mining and meta-modeling.
Optimus is a Process Integration and Design Optimization (PIDO) platform developed by Noesis Solutions. Noesis Solutions takes part in key research projects, such as PHAROS and MATRIX.
In machine learning, hyperparameter optimization or tuning is the problem of choosing a set of optimal hyperparameters for a learning algorithm. A hyperparameter is a parameter whose value is used to control the learning process.
