The progressive advancement in new technologies has introduced new types of equipment into the grid. Some of these devices, which have gained significant popularity in the last few decades are photovoltaic (PV) systems and wind power systems from the generation side, and energy-efficient lighting i.e. LED lamps and energy-efficient transportation via electric vehicles from the consumption side. All these equipment essentially consist of different kinds of power electronic converters with their associated control systems. There is always feasibility of mutual interactions between these non-linear power electronic devices with the grid as well as with other grid-connected equipment placed electrically close to each other, where the control loops can face new system dynamics. Adverse interactions can cause interferences and in the worst-case lead to instability issues. A mapping of the potential interactions, in the form of emission is needed to understand interferences at the device level and system level. The overall aim of this work is to enhance the existing knowledge about the power quality aspects of the different non-linear power electronic devices that are being increasingly connected to the grid especially in low voltage networks. The conclusions derived from this study can be extended to a broader scale where there is the feasibility of multiple non-linear power electronic-based devices operating together at medium and high voltage levels.

The main contributions of the work are classified into three main parts: 

In the first part, an in-depth study of interharmonics in PV systems is carried out. Different sets of field measurements and measurements from controlled but realistic laboratory environments are investigated for interharmonic existence, persistence, and propagation. To ensure the genuinity of the observed interharmonics and to address the different challenges associated with their estimation, combinations of methods are applied and results compared. The possible reasons for their origin are systematically established through a comprehensive study. The potential system impacts which it could create in the grid and to other grid-connected equipment are investigated. The possibility for aggregation of interharmonics when multiple sources are connected to the same point of common coupling is explored via a statistical approach.

In the second part, initially harmonic interactions in wind parks, and between PV and LED lamps, are first discussed. The time-varying harmonic interaction phenomenon is studied in detail with the help of a mathematical model as well as with the help of analysis of field measurements at multiple locations of a wind park. The outcome of this study contributes to the yet challenging problem of harmonic contribution estimation in twofold. (a) A method is developed from long-term field measurements with which one could potentially identify which source of emission dominates in the analysis period, and (b) Limitations of the extended mathematical model are identified and inferences from field measurements are linked to further improve the mathematical model. Further, some specific cases of harmonic interactions between PV and LED lamps are illustrated. These examples could be a guidance for power electronic designers to increase individual device immunity subjected to harmonic interferences.

Additionally in the same part, the impact of PV induced voltage variations on different topologies of LED lamps are investigated. The considered voltage variations are distinguished as overvoltage, undervoltage, rapid voltage changes, and voltage steps due to fast-moving cloud transients, due to inverter operation itself, and due to voltage regulations caused by load tap changing operations of distribution transformers. Due to PV induced voltage variations, LED lamps are impacted in various ways. LED lamps are either potential victims of these voltage variations or LED lamps to act as sources of increased grid distortions. As potential victims, the studied LED lamps have shown changes in the light output, instability issues, and degradation in the driver efficiency. As potential sources of grid distortion, LED lamps have exhibited increased harmonic and interharmonic emissions. The difference in impact has been linked to the topology of the lamps.

In the third part, the application of a deep leaning based unsupervised machine learning method to extract waveform distortion patterns in big data for enhancing power quality knowledge is illustrated. Specifically, signal processing of interharmonics with precise frequency and amplitude estimation needs the processing of data of large volume for a higher resolution. Thus, the interharmonics analysis in long-term measurements evidences the need for an automatic tool to assist the experts. In this work, a deep learning method for the identification of spectral patterns of time-varying waveform distortion in photovoltaic installations is proposed. The PQ big data with information on harmonic and/or interharmonics in PV installations is handled by a deep autoencoder followed by feature clustering. Measurements of voltage and current from four distinct PV installations are used to illustrate the method. The proposed method accelerates the process of manual interpretation and is a starting point to determine how to proceed further with the data analysis.

Continuous power quality monitoring allows grid stakeholders to obtain information about the performance of the network and costumer facilities. Moreover, the analysis of continuous monitoring allows researchers to obtain knowledge on power quality phenomena. Power quality measurements result in a large amount of data. Power quality data can be classified as big data, not only for its volume, but also for the other complexities: velocity, variety, and veracity. Manual analysis of power quality data is possible but time-consuming. Moreover, data reports based on standardized indexes and classical statistical techniques might hide important information of the time-varying behaviour in power quality measurements. Artificial intelligence plays a role in providing automatic tools for proper analytics of big data.  A subset of artificial intelligence called machine learning has enabled computers to learn without explicit programming. Driven by the huge improvements in computer processing, a subset of machine learning based on multiple layers of artificial neural networks has been developed to tackle increasingly complex problems. The so-called deep learning applications teach themselves to perform a specific task by automatically extracting essential features from the raw data. Despite the possibilities of automatic feature extraction, most applications of deep learning to power quality are still the same as expert systems or earliest machine learning tools. Moreover, most of the applications are based on synthetic generated data and supervised techniques. In this context, the main motivation of this thesis is providing a new tool based on unsupervised deep learning to handle analytics of time-varying power quality big data. 

The unsupervised deep learning method proposed in this thesis combines a deep autoencoder with clustering for extracting patterns and anomalies in power quality big data. The deep autoencoder maps the original data to a compressed format that contains the principal features of the data. Automatic results are provided by the deep learning, and inferences can be obtained without requiring prior knowledge of deep learning. The outputs from unsupervised deep learning can serve as a guide for further data analysis, highlighting important time steps within large power quality datasets. By following these indications from the deep learning results, experts gain valuable insights into power quality phenomena, which can be referred to as "learning from deep learning". The interpretation of the deep learning results in this thesis allowed to making proper inferences for patterns and anomalies. For power quality measurements synchronised with 24-h, the results allowed making inferences concerning daily variations, seasonality, and the origins of power quality disturbances. For power quality measurements non-synchronised with 24-h, the results could be interpreted visually through the distribution of the patterns in a physical variable, such as the dynamic operating conditions of an electrical railway power system.

An important contribution of this thesis concerns the physical interpretation of the phenomena is related to the anomalies in harmonics caused by geomagnetically induced currents. An interesting finding by applying the deep anomaly detection to measurements in the Swedish transmission grid is the damping of the anomalies caused by geomagnetically induced currents in the winter due to the heating load. This thesis also demonstrated that the signatures for anomalies in harmonic measurements in a Swedish transmission location are similar to the ones found in a low-latitude transmission location at the South Atlantic Anomaly due to geomagnetically induced currents. Moreover, by cross-checking the anomalies at the South Atlantic Anomaly with protection trips with undetermined causes, this thesis demonstrated that anomaly harmonics due to geomagnetically induced currents can cause protection mal trips.

This thesis demonstrates that unsupervised deep learning can serve as an additional tool for compressing time-varying power quality big data into a more interpretable form. Despite the application of an unsupervised method, power quality experts remain significant in power quality studies. The main conclusion of this thesis is that unsupervised deep learning enhances the understanding of power quality experts and provides a complementary approach for analysing and extracting insights from time-varying power quality big data.