Search

In many electronic and information engineering related disciplines, various complex machine learning and reinforcement learning optimization algorithms appear nowadays so as to handle huge amount of data related. Traditional optimization methods, which have to take the whole data space into consideration, cannot always perform well, and cannot always be completed in a short period of time as the data amount to be handled is huge. The object of the so-called optimization problem is to find the optimal solutions or quasi-optimal solutions in a complex and huge data searching space, and finding the efficient optimization algorithm which can handle data with characteristics of complexity, nonlinearity, and modeling difficulties for practical engineering, can be meaningful for electronic and information industries. Huber loss function is a typical traditional optimization method, with the error expressed via smooth mean absolute error (MAE) which has the advantages of MAE and mean square error (MSE) and uses MAE to enhance the robustness of MSE for those abnormal data points. However, Huber loss function employs gradient descent (GD) method to find the minimum value, which can easily be trapped into the local optimal solution. In this paper, a hybrid optimization algorithm, Huber loss function based on cockroach swarm algorithm with t-distribution has been proposed, it has better global convergence, and via numerical experiment, its efficiency in calculation has also been proved.

Although China's power grid construction is at the forefront of the world, due to the high degree of village dispersion and rugged terrain, some mountainous regions in southwestern China have not so far been connected to the state grid. A group of students in Tsinghua University founded a charitable project called ‘Dream Grid’ in 2014, aiming at building power systems that meet basic electricity demand of schools in remote areas. The project has been running for 4 years so far, attracting more than 80 students to participate in. More than 150,000 dollars has been raised by sponsorships and crowdfunding to support the construction of 60kW photovoltaic micro grids in 5 different areas. This paper takes one of these projects in Zamthang, Sichuan Province, China, as an example, and explores the design process of a hybrid solar-hydro microgrid for a rural primary school isolated from the national grid. Zongke Village Primary School, the electricity of which is usually supplied by small-capacity hydropower, is one of the largest village school in Zamthang County and has more than 100 students. However, the small capacity and lack of stability of hydropower makes it difficult to solve the electricity supply problem there. The power system design hopes to meet the basic daily teaching loads mainly with a solar system and the lighting demands at night by introducing a battery storage system. The original hydropower and one additional gasoline generator are considered as backup power when PV output is low and energy storage system cannot cover all loads. The size of solar panel and battery storage system is based on the load and insolation analysis. The power system operation is controlled by an energy storage bidirectional converter, which could dynamically adjust the energy distribution based on solar outputs and load power. The design plan which is composed of 10.5kW solar panels, 39kWh battery bank, 5kW gasoline generator and 20kW converter, is actually applied to the project in Zongke, and the entire system has been running for 12 months and produced more than 13MWh electricity so far. Economic analysis is also carried out in this paper, which shows that PV panels, batteries and converter makes up the most of total costs, and the system would yield a cost of $0.296/kWh over the 10-year lifespan. In addition, by providing electricity for the lighting of neighboring households at night and charging an appropriate fee, the cost of power generation at the school can be further reduced by approximately $0.042/kWh.