Electrical Engineering [Control Theory/Predictor]

Paper details

Subject: Engineering

Topic: Electrical Engineering [Control Theory/Predictor]

Kalman Filter applied to power budgeting of small satellites (cubeSat). This is a communications and control problem, where I'd like to have a controller developed (Kalman filter) to optimize power within a constellations of satellites. More importantly a Matlab script or Simulink diagram is desired for proof of concept. I have attached my proposal that has a few sources and I provided web url's that have information on the topic with ideas I'd like to use. The desired format is IEEE. Their website has a template. Abstract page is needed!

Extras ordered: Abstract page

10 PAGES WITH 15 SOURCES

Adaptive Radius-Directed Linear Kalman Filter Applied to Optimization of Power Budgeting in CubeSats

The adaptive radius-directed linear kalman filter (ARD-LKF) is a powerful predictor. This useful mathematical tool will be used as a predictor type controller implemented inside of a CubeSat. This controller will help optimize power consumption given the satellite’s limited battery capacity and small surface area. This will be applied to power budgeting CubeSat's in a constellation. The predictor will need to be designed so that it can extrapolate future positions of a desired satellite’s location given it’s process is nonstationary i.e. mean and variance are changing in time. Using this ARD-LKF, the satellite can be placed in optimal positioning for transmitting and receiving signals, such that minimal power is lost in the medium. A limitation in a classic kalman filter is assuming priori knowledge of model parameters. Given this application, having complete priori knowledge is not possible, so an adaptive filter will be used. An adaptive filter will be able to take into account changing model parameters and estimate them together with the dynamic state. This can be implemented in multiple ways, but for low complexity and practical implementation the covariance matching algorithm will be used.

The range in which the Kalman filter will control will be dictated by an inverse distance weighted interpolation. This means a particular satellite will be controlled with weighted filters that become more influential in relation to nearby satellites and less influential to satellites that are far away. This will be applied to the kalman filter by having the calculated desired radius used as a measurement to be fed into the kalman update part of the prediction.

The main focus of this study will be to take into account fluctuations and perturbations, that are common for a CubeSat to encounter in orbit, and be able to predict where a satellite’s location will be given these random variables. It’s important to note that since the system is a constellation of CubeSat’s, the position of the desired satellite in relation to others satellites is what will be predicted using the controller, and not the satellites position relative to a transmitting or receiving point on earth.

This same concept can be applied to the constellation in regards to minimize lag. This will aid optimal power budgeting by selecting the satellite's with lowest number of running tasks to send/receive messages. Due to the randomness of variations between program workloads and when a user uplinks these commands; very little can be done with respect to forecasting and predicting how a specific task will affect the system, therefor the study of lag time analysis will be omitted.

Sources:

“Random Processes for Engineers” Dr. Snider

“Power Consumption.” Power Consumption | Raspberry Pi Dramble, www.pidramble.com/wiki/benchmarks/power-consumption.

Arnold, Scott Sterling, et al. “Energy Budgeting for CubeSats with an Integrated FPG.” doi: 978-1-4577-0557

Albertella, Alberta & Migliaccio, Federica & Reguzzoni, Mirko & Sansò, Fernando. (2004). Wiener filters and collocation in satellite gradiometry.

“Least Mean Square (LMS) Adaptive Filter.” Least Mean Square (LMS) Adaptive Filter - National Instruments, 23 Aug. 2013, www.ni.com/example/31220/en/.

J. Zhou and Y. Zhu, "The Linear Minimum Mean-Square Error Estimation with Constraints and Its Applications," 2006 International Conference on Computational Intelligence and Security, Guangzhou, 2006, pp. 1801-1804.

doi: 10.1109/ICCIAS.2006.295373

“Means of Calculating Low-Orbit Satellite Visibility Statistics .” Itu.int, www.itu.int/dms_pub/itu-r/opb/rep/R-REP-SA.2066-2006-PDF-E.pdf.

Haag, Michael. “Stationary and Nonstationary Random Processes .” OpenStax CNX, 18 July 2005, cnx.org/content/m10684/2.2/.

Huo, Lei, and Zhiliang Wang. “ A Target Tracking Algorithm Using Grey Model Predicting Kalman Filter in Wireless Sensor Networks.” IEEE Conference Publication, IEEE, 1 Feb. 2018, ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8276814.

Zhang, Qun, et al. “Noise Adaptive Kalman Filter for Joint Polarization Tracking and Channel Equalization Using Cascaded Covariance Matching.” IEEE Journals & Magazine, IEEE, 23 Jan. 2018, ieeexplore.ieee.org/document/8267189/.

Web url’s for reference.