Octavian Cuibus

Controller synthesis method for Discrete Event Systems

Many applications contain plants that are Discrete Events Systems (DES). They have to be controlled such that DES fulfills some specifications like: avoid the deadlocks, reach or avoid the reaching of some given states, execute or avoid the execution of some given sequences of events, execute cyclically sequences of events with the shortest periods, etc. In the current study the plants are modeled by Delay Time Petri Nets (DTPN) and the controllers are Time Petri Nets (TPN) models. The controllers can be described by a particular Time Petri Net Language (TPNL). The TPNL descriptions can be transformed into Lisp descriptions. The latest are used by a Genetic Programming (GP) method for the controller synthesis such that DES meets most accurately the system requirements.

Automatic Linear Robot Control Synthesis Using Genetic Programming

An automatic controller synthesis method for a single axe linear robot is considered. The robot motions are modeled by a Delay Time Petri Net (DTPN). The search refers to automatically finding a controller modeled by a Time Petri Net (TPN) that fulfills some specified requirements. The controller model is synthesized using a Genetic Programming (GP) method. The mapping between TPN model and the tree representation of individual genotypes is performed using a formal language named here TPNL (Time Petri Net based Language). This language is suited for formal description of the controller behavior traits like sequential, concurrent, selection, loop or input/output. The use of control traits guaranties the construction of individuals that are capable and useful to control the robot moves. To diminish the search durations, besides the usual genetic operators like mutation, permutation and crossover, a new atrophy operator was introduced.

Implicit cooperative control achieved by generic genetic fuzzy logic

The paper presents a cooperative control approach applied to a lake system that is implemented through a set of communicating agents. The distributed controllers use fuzzy logic rules to get the control decisions. To improve the local control performances the agents send each other their contexts. The control improving involves more terms in the fuzzy logic control rules. This has the effect of strongly increasing the number of rules. Hierarchical architectures are chosen to diminish the rule numbers of different methods. To obtain the unknown fuzzy logic control rules the generic genetic approach is used. The control rule sets, the simulations results and their comparison are given for some cooperative methods.

Genetic programming synthesis of discrete event controllers applied to urban vehicle traffic control

The paper presents a new method for generating discrete event systems as control units for a type of plants which can be modeled as delay time Petri nets. The control unit contains many transitions joined together by a set of operands and it is generated by means of the genetic programming method using a Lisp representation of the solution. Transitions are characterized by an enabling condition (reaction or feedback) and an effect (control), which represent the interaction with the plant. Crossover and mutation operators are defined for Lisp expressions. The method is applied to urban vehicle traffic control.

Price based negotiation strategy for urban vehicle traffic control

The paper presents a new explicit cooperative control strategy in which agents responsible for certain parts of the system formulate requests to their neighboring agents and carry negotiations regarding the cost of each request. The interlocutory agents respond with the demanded revenue for the request and a new request proposal. The negotiation algorithm describes the behavior of agents during each phase of the negotiation. Three different methods for calculating cost and revenue functions are given. The novel control strategy is applied to urban vehicle traffic (UVT) control.

Cooperative Control for Urban Vehicle Traffic

Abstract The paper presents a cooperative control approach applied to an urban vehicle traffic system that is implemented through a set of communicating agents. The distributed controllers use fuzzy logic rules to get the control decisions. To improve the local control performances the agents send each other their contexts. Improving the control involves more terms in the fuzzy logic control rules. This has the effect of strongly increasing the number of rules. Hierarchical architectures are chosen to diminish the number of logic rules for different methods. To obtain the unknown fuzzy logic control rules the generic genetic approach is used. The control rule sets, the simulations results and their comparison are given for some cooperative methods.

Cooperative Control of Hydro-Power Systems

Abstract The schedule and control of a set (chain) of hydro-power plants is taken into account. Even if the water input flows vary significantly and the estimations often fail, the lake levels have to be constrained and the power generation should be preferentially provided during the peak hours. The proposed solution for long time schedules of the generator working times are based on genetic algorithm. Due to the fact that the long time estimated input flows are different of the real input flows, the off-line scheduling performance is low, so cooperative control methods that works on-line are used to improve the schedules for each quarter of a day. Fuzzy logic rules are used for the short time scheduling and genetic algorithms are used off-line for obtaining the rules that work on-line. The cooperation of neighbor controllers is necessary to avoid the conflicts involved by their local objectives.

Automatic Control Synthesis of Hydro-Power Systems

Abstract The control of a set of hydro-power plants with or without reservoirs is taken into account. Even if the water input flows vary significantly, the lake levels have to be constrained and the power generation should be preferentially provided during the demanded peak hours. Each power station has assigned a local controller capable to communicate with its neighbor controllers. The proposed genetic programming method automatically generates a distributed Time Petri Net (TPN) that is capable to minimize the lake level variations and to update the decisions to start the generator(s) with the input flow levels. The cooperation of neighbor controllers improves the local control decisions taking into account the neighbor decisions and so reducing the daily generator start number.