Using the hierarchical temporal memory spatial pooler for short-term forecasting of electrical load time series

Abstract

In this paper, an emerging state-of-the-art machine intelligence technique called the Hierarchical Temporal Memory (HTM) is applied to the task of short-term load forecasting (STLF). A HTM Spatial Pooler (HTM-SP) stage is used to continually form sparse distributed representations (SDRs) from a univariate load time series data, a temporal aggregator is used to transform the SDRs into a sequential bivariate representation space and an overlap classifier makes temporal classifications from the bivariate SDRs through time. The comparative performance of HTM on several daily electrical load time series data including the Eunite competition dataset and the Polish power system dataset from 2002 to 2004 are presented. The robustness performance of HTM is also further validated using hourly load data from three more recent electricity markets. The results obtained from experimenting with the Eunite and Polish dataset indicated that HTM will perform better than the existing techniques reported in the literature. In general, the robustness test also shows that the error distribution performance of the proposed HTM technique is positively skewed for most of the years considered and with kurtosis values mostly lower than a base value of 3 indicating a reasonable level of outlier rejections.

Open Power System Data – Frictionless data for electricity system modelling

“…It provides a wide range of data at one place, is easily accessible, clean and ready-to-use, permanently available and versioncontrolled. The large number of users -around 100, 000 unique visitors during 2017) -and, more importantly, the amount of research that makes use of OPSD -26 published papers by the time of writing since the go-live in late 2016, out of which 12 are published papers in high quality journals indexed in the SCI/SSCI [13,[40][41][42][43][44][45][46][47][48][49][50] and 14 in other journals, conference papers, books and grey literature [51][52][53][54][55][56][57][58][59][60][61][62][63][64] -suggest that the platform fulfils a need and provides value to electricity system modellers.…”

Anomaly Detection for IoT Time-Series Data: A Survey

“…A recent development within the Artificial Neural Network domain is a process described as Hierarchical Temporal Memory (HTM) [78], this process is a bio-inspired model for processing time-series based upon the behaviours of the Neocortex. This method is applied to sequential streamed univariate data in [79], [80] and compared against a range of predictive models for time-series modelling. The technique is further applied to the anomaly detection problem in [81], [82], [83], [84] of note is the noise resistance of the approach as well as the ability for continual online learning allowing for the method to adjust to changes in data distribution over time without extensive off-line retraining.…”

A New Hierarchical Temporal Memory Algorithm Based on Activation Intensity

“…Osegi [ 23 ] applied HTM into the task of short-term load forecasting using spatial pooler and a temporal aggregator, which transform SDRs into a sequential bivariate representation and makes temporal classifications from the SDRs. They verified that HTM has stronger noise resistance and can outperform most existing artificial intelligence neural technologies in short-term load forecasting tasks.…”