November 12, 2019 by Sam Holland
As much of the world transitions from fossil fuels to green energy, a new, transformed transmission system must be realised. Indeed, electrical engineers must ultimately transform the traditional grid into a smart grid.
In the past, the transmission line infrastructure was not capable of transmitting voltages at frequencies higher than 50 to 100Hz. It was such a low range that, as the transmission distances inevitably increased, the loss of power on the line increased with it—leading to marked inefficiency rates and a loss in performance for end utility operations. After all, the electrical cables of such traditional systems are not capable of handling high-frequency (HF) transmissions (consider the energy radiation and reflections that occur at discontinuities), and this prevents the signal from reaching the intended destination.
A close-up of a transmission system. Image Credit: Bigstock.
Accordingly, the modern solution is to ensure an infrastructure that accommodates a high-frequency voltage (above 30kHz) transmission, without significant losses along the line. Waveguides, optical transmission, and of course, smart grids are some of the latest milestones in the field of electrical engineering: their efficiency can ensure the ideal HF voltage transmission with minimum power loss.
That said, transmitting the desired HF voltage signal is still far from simple. Communication channel equalisation, in addition to the optimisation of the transmission lines, will be required to bring about the automatic transmission of HF voltage signals. The channel equalisation involves withdrawing both the channel effect and the noise that may interfere with the original signal. Once the signal is optimised, filtering methods are then to be utilised to filter out the unwanted components, in order to extract the desired signal at the destination in question.
Such a general process of HF voltage transmission can be implemented in different ways, and these involve various filtering and equalisation techniques. It was in a study published by the European Journal of Electrical Engineering that a team of researchers from the Guiyang Bureau of UHV Transmission Company worked to optimise HF voltage signals, through remote control, to achieve better stability in operating power systems.
While the facilitation of high voltage transmission has been traditionally expedited using Fast Fourier Filtering (aka FFT), adaptive filtering, notch filtering, and so on, it’s the following methods that appear to offer a better solution to the problem at hand.
An adaptive filter is a system with a linear filter: one that has a transfer function controlled by variable parameters, as well as a means to adjust those parameters according to an optimisation algorithm. This filtering technique (exemplified in the below diagram) is primarily used to subdue the interference that is encountered in HF voltage transmission and reception.
A diagram that exemplifies an adaptive filtering system. This employs least mean squares. Image Credit: Wikimedia Commons.
The underlying principle of this process makes use of mean squared error criteria to adjust the incoming dynamic parameters. This way, the filter produces estimates of the incoming voltage signal’s useful component and eliminates any errors that it encounters in the process. The filter tracks the changes in the input signal and then employs the least mean squares (LMS) algorithm to adjust the parameters to augment the output signal.
However, just like any other filtering technique, this method is not perfect. The LMS algorithm has a slow convergence rate, and while users can at least achieve a speedy convergence by increasing the step size, this, in turn, will inevitably go wrong and ultimately expand the steady-state value of the mean squared error.
Suffice to say, there is a trade-off involved, and even if users choose to overcome this—namely by increasing the simulation number—this will then make it difficult to select the right parameters for the filter, given that the frequency tends to change due to multiple narrowband interferences.
Another way to employ adaptive filtering is with wavelet transform. This allows users to overcome the said narrowband interference that users would encounter if they applied the above-mentioned technique. Similarly to the previous method, this technique also makes use of the LMS and works by analysing the signal at partial discharge levels before carrying out the relevant processing. This way, not only is the narrowband interference problem resolved, but the stability of the system is improved, too.
Although the above does deliver the desired results, as is so often the case, it comes at the cost of high computational power.
The Gaussian filter, unlike those previously mentioned, is non-linear. Per its namesake, the impulse function of such a filter is called a ‘Gaussian function’. This technique, in conjunction with its capacity to clamp down on interference, is also deployed for the automatic transmission of HF voltage signals, via remote control.
Both an improved noise cancellation percentage and the overall stability of the system is achieved when utilising this filtering technique.
Power spectrum analysis and the use of the matched filter in this interference suppression technique each improve the transmission at high voltage, namely by cutting down on the noise and improving the output signal stationarity. The corresponding algorithm, however, is not able to feed back such efficient results in real time: this makes it an impractical option for use on a large scale.
For the said techniques to work and improve the transmission of HF voltage, there are further factors that have to be taken into consideration. The filtering techniques must be fashioned in a manner that ensures the bit error rate is kept to a minimum, whereas the signal-to-noise ratio (both of which are defined below) is kept to a desired maximum.
To achieve the above, the voltage signal and the channel of transmission need to be modelled initially; then, interference filtering is required to cancel out the noise contamination—all before the channel is finally realised for optimal HF voltage transmission.
Bit error rate is the ratio of error bits to the signal bits received at the receiving end of the transmitted signal. The ratio signifies the quality of the transmission of a digital device. The value of this ratio being low is something that is desired in a received signal when an HF voltage transmission is concerned.
In this context, signal-to-noise ratio, or SNR (exemplified in the below graph), refers to the level of the signal compared to that of the noise encountered in the received signal. A high value for this ratio is desirable in the transmitted signal, and accordingly, this is another performance parameter for the corresponding filter design.
A graph displaying measured in decibels (dB), a signal-to-noise ratio (SNR) in a received signal. Image Credit: MathWorks.
Despite the current methods for the high-frequency transmission of voltage under remote control, further work is still needed to enhance the performance of the relevant filters, and of course, to ensure that the right parameters are met.
Indeed, given the increasing demand for energy and long-distance transmission, it’s time for electrical engineers to pull out all the stops—to ultimately render the required solutions a reality.
