| The development of the solar dish/Stirling system(SDSS),as a promising green power generation technology,has received great interest from researchers and governments.China,as the world’s pioneering country in green energy production,has heavily invested to develop solar energy technologies in the recent past.Nevertheless,commercialization and characterization of dish/Stirling technologies need a deep understanding of the influence of different design,opt-geometrical,and operational parameters on their performance optimization.The literature survey revealed a lack of parametric studies on dish/Stirling systems.Moreover,there is a remarkable absence of multi-objective optimization investigations during the design stage of the SDSS,particularly for the concentrator and receiver.Therefore,the research presented in this thesis will bridge these gaps and develops a novel technique of a reliable advanced optimization method to deeply determining the optimal design parameters and investigating the performance optimization of the SDSS.And meanwhile,design a commercial scale SDSS with improved components for power generation applications.Hence the primary objective of the research presented in this thesis is carrying out a detailed theoretical model implemented in MATLAB software for calculating the electrical power and the overall efficiency of the SDSS based on the opt-geometric sizing and thermodynamic analyses of SDSS.A comprehensive parametric analysis is conducted to identify the impact of the receiver operational temperature,the concentrator diameter,the wind speed,the direct solar radiation,and the ambient temperature not only on the net power and overall efficiency of the SDSS but also the thermal energy losses of the receiver.Thereafter,an optimization algorithm has been also developed for simultaneously maximizing the output power and total efficiency of the SDSS using the Multi-Objective Particle Swarm Optimization(MOPSO).Nine influential decision variables have been considered and the optimal Pareto frontier was determined as well as the optimal solution was selected by applying decision-making approaches.The results showed that the proposed MOPSO is a feasible approach to achieve a maximal power output of 23.46 k W with an optimal final total efficiency of 30.15%that yielded a very low deviation index from the ideal solution if the SDSS is designed with higher values of design parameters.In addition,a sensitivity analysis was also carried out to identify the influence of the dish concentrator diameter,the rim angle,the mirror soiling coefficient of the concentrator,and the design concentration ratio on the SDSS performance optimization.A series of Pareto optimal frontier and important quantitative results were obtained to allow designers of dish/Stirling systems to incorporate the effect of design parameters on the system performance.On this basis,a new commercial SDSS with a rated power of 25 k W was effectively designed and experimentally established at Tianjin,China based on the obtained results.The design criteria,modular component’s technical selection,structure design,and experimental implementation steps of the Tianjin solar dish-Stirling system(TJ-SDSS)were described as well.The system site is characterized by a northeastern China climate.The tests to be carried out on this system are particularly motivating because there are few demonstrative dish-Stirling systems established in China’s Jingjinji Metropolitan region as there is no evidence of the influences of the particular weather of this region on the feasibility and performance operation of this kind of CSP technology.Furthermore,a dynamic thermo-economic model was also developed to dynamically simulate the proposed TJ-SDSS operation,in order to analyze its performance,under conditions of Tianjin,China.An analysis of the daily,monthly,and annual performances of the TJ-SDSS was performed.In addition,techno-economic viability for assessing the electricity cost of the TJ-SDSS was also performed.The results indicated that the proposed TJ-SDSS produces 28.748 MWh annually,with a yearly net overall efficiency of 19.55%,and achieves maximum energy of 3.8 MWh with a monthly average efficiency of 22.75%in June.From the energy cost analysis,the electricity cost of the TJ-SDSS is estimated to be 0.2565$/k Wh～(1.719CNY/k Wh).In conclusion,the present study identifies the economic viability of the dish Stirling technology in China,taking into account China’s feed-in tariff of CSP systems(1.2 CNY/k Wh),and it also provides new theoretical guidance for predicting the annual performance and evaluating the economic situation of the dish/Stirling systems.In another separate context,this thesis additionally provides insight into the innovations and application of hybrid artificial intelligence models for predicting the dynamic performance of sustainable energy systems with complex interactions such as the complicated dish/Stirling technologies.Hence,several novel optimized artificial intelligence approaches were developed for modeling different cases of dish/Stirling technologies under various operating conditions.In the first case study,the performance of the solar dish/Stirling power plant(SDSPP)was predicted using the random vector functional link(RVFL)network as a novel artificial intelligence framework with superior advantages.Moreover,four types of novel metaheuristic algorithms were integrated with the original RVFL namely;Spherical Search Optimization(SSO),Chimp Optimization Algorithm)CHOA),Whale Optimization Algorithm(WOA),and Particle Swarm Optimization(PSO)to generate a fine-tuned prediction model and identify the optimal RVFL parameters.The five optimized RVFL models including RVFL-CHOA,RVFL-PSO,RVFL-WOA,RVFL-SSO,and single RVFL were compared and applied to predict the monthly power production and the instantaneous electrical power of a commercial SDSPP.A comparative statistical analysis on the performance of five models was conducted towards identifying the more accurate model for SDDPP performance prediction.The statistical findings prove that the R~2value of the RVFL-CHOA model ranges between 0.9180-0.9992 for the testing phase;which is higher than that of RVFL-PSO,RVFL-WOA,RVFL-SSO,and single RVFL which ranges between,0.8409-0.9799,0.8518-0.9899,0.8698-0.9799,and 0.7239-0.9798 for the testing phase,respectively.In the second case study,another innovative hybrid method is developed to predict the thermal performance parameters of solar dish collector operating with multi-walled carbon nanotube/thermal oil nanofluid using a modified algorithm of the adaptive neuro-fuzzy inference system(ANFIS)integrated with equilibrium optimizer(EO).To evaluate the performance of the developed method,ANFIS-EO is compared with a single ANFIS and ANN.The statistical results revealed that the ANFIS-EO method had the efficient prediction accuracy compared with ANFIS and ANN models.In general,this thesis presents an emerging comprehensive approach for optimizing,designing,and constructing a solar dish/Stirling power system and analyzing its dynamic thermo-economic performance.Moreover,the findings of this work can be used for exploring the maximum performance of each examined concentrator configuration at any geographic location.In addition,the developed hybrid artificial intelligence models that were developed in this thesis can be regarded a promising prediction optimization tools and constitute a step forward in the applications of dish/Stirling systems by proposing that innovative,hybrid,and perhaps more complex forecasting models,which are coherently more precise than classical neural models.It would be stated that each objective of the research presented in this thesis has been accepted for publication via six high quality published articles;namely,(one review article and five research articles),which this,in turn,reflects the great scientific contributions that this thesis made to the research community interested in solar dish/Stirling systems. |
PDF Full Text Request