Home

PROJECT OBJECTIVE

BACKGROUND

click image to enlarge

          The primary objective of this research project was to optimize a prototype ground-based cloud detection device, used to calculate cloud velocity. The first prototype was built by the Kleissl Solar Resource Assessment and Forecasting Laboratory at the University of California, San Diego under the guidance of the associate director of UCSD’s Center of Energy Research (CER), Dr. Jan Kleissl. Above is a an image and a CAD assembly of the optimized product. It was created by scaling down the size of the original sensing array and creating a higher quality cloud speed and direction sensing system. This includes having higher sampling rates, creating a weatherized housing enclosure, and optimizing the post processing algorithm of data (onboard) before relaying measurements to a centralized computer. The design and fabrication of the new and improved prototype was conducted by the collaboration of four senior level undergraduate engineering students: Andy Chen, Jeff Head, Tyler Capps, and Victor Fung. For more detailed information please go to the Final Design page or the Executive Summary.

I

HOW IT WORKS

            The colossal consumption rate of limited fossil fuel creates significant problems for the environment and [available] energy resources. To ensure that there will be a sufficient amount of energy to continuously drive the world’s economy and food supply for the present and the future, the [development and] transition to a more abundant renewable source of energy becomes necessary. The challenges of converting to a new energy resource should be tackled today in order to prepare for the eminent depletion of fossil fuel.

            Annually, the Earth receives approximately 8,000 times more solar energy than the amount of energy consumed globally, proving solar power to be a reliable candidate as the next primary energy supplier. However, the unpredictability of its variable nature due to different cloud coverage stands as a large scale engineering obstacle that hinders the prominent widespread penetration of solar power into the energy market. Therefore, it is essential to accurately forecast available solar resources through the tracking of cloud movement, which can be accomplished with a grid of ambient light sensors to sense a cloud’s shadow as it sequentially shade different sensors.

            The figure above is an image of how the system would be integrated into a solar power plant. With calculated cloud velocity and direction, and a known distance to the solar plant solar companies can accurately predict when solar panels will be under cloud coverage.

SUMMARY OF PERFORMANCE RESULTS

            The performance of the final design for the sensor array was judged mainly on the accuracy of cloud velocity measurements over a long period of time (at least one hour). Due to the inherent uncertainty of cloud motion, it was difficult to obtain a correct, quantified number for comparison. Therefore, the accuracy of the system was defined by comparing the velocity measurements of our device with those calculated using post processing on excel. The results were then compared to the known values, and it can be seen below that calculated values agreed.

             The process of measuring cloud speed and direction is rather intuitive. The device consists of 15 ambient light sensors centered about 1 central sensor. The outer sensors in the array will be the first to detect cloud coverage as the cloud’s shadow passes over them. This detection will be represented by a decrease in measured solar irradiance. Following this, a time lag will ensue. This time lag is the time it takes the cloud to travel from the outer senor to the center sensor. With the measured time lag and a known distance between the pyranometers, the velocity can be calculated. As for determining the direction of the cloud, this is calculated using the correlation coefficients between the central and outer sensors. The most correlated pair, which describes the similarity between the measurements, identifies the pair of sensors most aligned with the cloud’s direction.

            Above is a MATLAB simulation of the research project. A (red) cloud, modeled as a circular object for simplicity, is translated over the semi-circle array of sensors (green). From here, it is easy to observe a time-lag comparing any of the 15 (outer) sensors to the central sensor. With a measurable time-lag, and a known distance between the outer sensor and central sensor, it is feasible to compute the speed of the shadow, which corresponds to the speed of the cloud.