TY - THES
AU - Miniesy, Mohammed Samir Mohammed
PY - 1971
TI - Application of modern control techniques to power systems
KW - Thesis/Dissertation
LA - eng
M3 - Text
AB - A power system may be subjected to different types of disturbances. The control strategy to be taken in order to preserve system stability depends on the severity of the disturbance.
For very severe disturbances, power system stability can be improved by sudden changes in the electric power network such as the insertion of braking resistors, generator dropping or load shedding. A unified treatment of optimum switching is presented by considering the switching instants to be elements of a generalized control vector. Dynamic optimization is then applied to determine optimum switching instants.
Less severe disturbances can be overcome by employing governor and/or voltage regulator controls. The governor control problem for a large signal model of interconnected power plants is investigated via the multi-level concept. A two-level controller for interconnected power plants is discussed. Each plant has a first-level local optimal or suboptimal controller. The second level of control is an intervention control performed by a central co-ordinator. If a sudden system disturbance causes the system angular acceleration to exceed preset tolerances, a priority interrupt to the central co-ordinator initiates intervention control. Angular velocity deviations of all plants are transmitted to the co-ordinator. This data is used to generate coefficient data for each plant. On receiving its coefficient data, each plant generates a local second-level intervention control which augments first-level local control.
The Load-Frequency Control problem, due to minor or routine disturbances
caused by load changes, is investigated. Since the incremental power demand in a power system is not always known a priori, direct application of the optimum linear-state regulator to Load-Frequency Control is not possible. Furthermore, Load-Frequency Control generally requires the use of an integral-type control operation to meet the system operating specifications. This requirement is introduced into the formulation of the optimum Load-Frequency Control problem presented in this thesis.
Two methods are suggested for demand identification. The first method makes use of differential approximation. The second method makes use of a Luenberger observer to identify unmeasured states. The optimum control is a linear function of measured states, identified unmeasured states, and the identified incremental power demand.
A method is given for solving, suboptimally, the problem of optimum-load frequency sampled-data control with either unknown deterministic power demand or randomly varying system disturbances. It is shown how to modify an optimum continuous control to obtain optimum control in the case of discrete-data transmission and unknown deterministic demand.
The case of random power demand and random disturbances is treated by introducing an adaptive observer. A three stage systematic design procedure is given. The effectiveness of Load-Frequency Control using an adaptive observer is illustrated by an example.
N2 - A power system may be subjected to different types of disturbances. The control strategy to be taken in order to preserve system stability depends on the severity of the disturbance.
For very severe disturbances, power system stability can be improved by sudden changes in the electric power network such as the insertion of braking resistors, generator dropping or load shedding. A unified treatment of optimum switching is presented by considering the switching instants to be elements of a generalized control vector. Dynamic optimization is then applied to determine optimum switching instants.
Less severe disturbances can be overcome by employing governor and/or voltage regulator controls. The governor control problem for a large signal model of interconnected power plants is investigated via the multi-level concept. A two-level controller for interconnected power plants is discussed. Each plant has a first-level local optimal or suboptimal controller. The second level of control is an intervention control performed by a central co-ordinator. If a sudden system disturbance causes the system angular acceleration to exceed preset tolerances, a priority interrupt to the central co-ordinator initiates intervention control. Angular velocity deviations of all plants are transmitted to the co-ordinator. This data is used to generate coefficient data for each plant. On receiving its coefficient data, each plant generates a local second-level intervention control which augments first-level local control.
The Load-Frequency Control problem, due to minor or routine disturbances
caused by load changes, is investigated. Since the incremental power demand in a power system is not always known a priori, direct application of the optimum linear-state regulator to Load-Frequency Control is not possible. Furthermore, Load-Frequency Control generally requires the use of an integral-type control operation to meet the system operating specifications. This requirement is introduced into the formulation of the optimum Load-Frequency Control problem presented in this thesis.
Two methods are suggested for demand identification. The first method makes use of differential approximation. The second method makes use of a Luenberger observer to identify unmeasured states. The optimum control is a linear function of measured states, identified unmeasured states, and the identified incremental power demand.
A method is given for solving, suboptimally, the problem of optimum-load frequency sampled-data control with either unknown deterministic power demand or randomly varying system disturbances. It is shown how to modify an optimum continuous control to obtain optimum control in the case of discrete-data transmission and unknown deterministic demand.
The case of random power demand and random disturbances is treated by introducing an adaptive observer. A three stage systematic design procedure is given. The effectiveness of Load-Frequency Control using an adaptive observer is illustrated by an example.
UR - https://open.library.ubc.ca/collections/831/items/1.0093304
ER - End of Reference