Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.781322
Title: Modelling and control of biomass and photovoltaic distributed generation systems
Author: Rufa'I, Nabila Ahmed
ISNI:       0000 0004 7966 9507
Awarding Body: University of Leeds
Current Institution: University of Leeds
Date of Award: 2019
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Abstract:
Distributed generation units using various renewable energy sources are rapidly infiltrating traditional electric power networks. They bring the benefits of hugely increasing worldwide energy supplies and reducing greenhouse gas emissions. Several distributed generators (DGs) can be integrated to form a microgrid (MG) to supply the local power needs, hence curbing power transmission congestion and delivery costs. A MG can manage the transfer of electricity between regions, and improve the robustness of the electrical network and its resilience in times of disastrous events. However, renewable sourced DGs using voltage source inverters with LCL filters are sensitive to non-ideal grid conditions such as impedance variation, harmonic corruption, voltage dips or swells and imbalance. Control techniques capable of working against all abnormal grid conditions, hence ensuring robust and stable power flow in the network, are crucial. One of the main contributions of this work is the development of a new adaptive notch filter for the control of grid connected DGs. Grid-side LCL power filters in the converter output, using small capacitors and inductors, are effective in attenuating the PWM carrier and side-band voltage harmonics, and are widely applied. However, they are characterised by a unique resonant frequency, at which the gain of the input current is magnified. This frequency must be removed from the converter output, and therefore from its current control loop. A notch filter (NF) in the DGs' current control loop can actively curb this resonant frequency problem, but only at the specific value determined by grid impedance and LCL parameters. With a dynamic grid having variable impedance and background harmonics, the NF parameters need to be adaptive. The new method detects the changes of the resonant frequency by analysing the grid-side measured current on-line in real-time using the fast Fourier transform (FFT) technique, and the resultant frequency value is then applied to adjust the parameters of the notch filter. The method is tested and found to be able to track the changes of resonant frequency quickly and accurately, hence allowing the converter to give high performance current control. For testing and validating the control schemes, a DG system with two or more renewable sourced generators is required. A micro-grid consisting of a biomass generator and PV-plant jointly supplying the local load is therefore built in simulation. Detailed models for the biomass-based thermal energy conversion plant driving a synchronous generator and a PV DG are developed. The robustness of the biomass DG (BDG) controller is investigated under load perturbations, and the relationship between the supply system and the inertia of the BDG system is analysed. The study includes the negative effects of unbalanced loads, and a proposed solution is to compensate the negative sequence component of the unbalanced load using the advanced decoupled double synchronous reference frame (DDSRF) technique. Another contribution lies in the development of a new Optimised Flexible Power Control (OFPC) scheme for reference current generation for PV DGs operating under abnormal grid conditions. DGs are required to supply active and reactive power without tripping under low or unbalanced voltages, but suffer from high power ripples due to the effects of negative sequence current and voltage. Treating the suppression of real and reactive ripples under a specific unbalanced voltage condition as a constrained multi-objective optimisation problem, the proposed OFPC defines a cost function as the sum of weighted active and reactive power ripples with a control variable. A genetic algorithm is applied to search for the optimal variable value. With this scheme, the reference current for the DG current controller gives good power control, and the results compare favourably with conventional methods.
Supervisor: Zhang, Li ; Chong, Benjamin Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.781322  DOI: Not available
Share: