Tony Yip

Active Control of Power Flow in Distribution Network Using Flexible Tie Switches

To accommodate high penetration of distributed generation (DG), the distribution network is moving towards the direction of active control and active management. This paper presents a new type of flexible tie switch (FTS), installed in place of traditional sectionalizers or tie switches, which can be applied for active power flow control in a distribution network. FTS can potentially optimize the existing network capacity so as to accommodate more DGs. In order to analyze its effectiveness, the simplified mathematical model of FTS in dq coordinates was first derived. The dynamic power control performance of FTS was studied in PSCAD with typical dual closed-loops control algorithm adopted. Minimum network losses and voltage profile optimization were set as the goal of power flow control. Multi-objective optimization function which takes FTS as control variable was built, considering voltage deviation constraint, capacity limitation of FTS and its power losses. Hybrid particle swarm algorithm was used to search for optimal solution under different load level conditions. Results show that the network losses are reduced significantly and the voltage profile is also optimized, when FTSs are used for power flow control in a distribution network.

A new secure method for the transmission of protection information between substations based on compressed sensing

Wide-Area protection uses the global information to detect faults and to achieve stability control, which is a hot research topic nowadays. However, wide-area protection system needs to use information from a number of substations. Therefore, communications reliability requirement is high. To ensure reliability of the information transmitted between substations, this paper presents an encoding method based on compressed sensing (CS). This method creates sparsity for the data to be transferred and maps the information to the frequency domain before transmission. At the receiving end, an orthogonal matching pursuit (OMP) reconstruction algorithm is used to reconstruct the original information. The simulation results show that even with the worst case bit change, or when the communication channel has noise, the normalized root mean square error (NMRSE) is still acceptable. The method simplifies the encoding and decoding complexity, greatly improves efficiency.

Wide Area Power System Fault Detection Using Compressed Sensing to Reduce the WAN Data Traffic

With the increasingly complex power system, wide area protection, using global data obtained from different substations through communications, has been a hot research topic for some time. However, the overall transmission of large amounts of data will cause communication network congestion, which will lead to delay and loss of data. Therefore building an algorithm which can make use of a reduced number of global data to identify the fault area is very useful. This paper proposes a down-sampling matrix to reduce the original data. For example, a protection system requiring 240 feature points of voltage data, if using the down-sampling matrix, will need only a minimum of 24 points, and still has a high probability to identify the fault zone. Simulation results show that when the data size M > 0.3, the result of classifying adjacent bus fault point is credible (greater than 60%), and when the data size M > 0.05, the result of classifying the non-adjacent bus fault point is credible (greater than 72%).

A hierarchical optimization strategy for distribution network reconfiguration considering the access of flexible interconnect devices

To achieve active control and active management of distribution network, flexible interconnect devices (FID) based on power electronics are proposed to optimize the power flow. This kind of device can be a good supplement to conventional control means due to its ability to regulate the power flow. This paper addresses network reconfiguration with the application of FID for improved flexibility in distribution network operation. In particular, a hierarchical optimization strategy based on hybrid coding is proposed specifically to solve the optimal problem of network reconfiguration with FID. A case study based on IEEE 33-bus system was carried out for strategy verification. The results show that it is possible to realize the better performance of distribution system by the collaborative optimization of network reconfiguration with FID.

A two-stage random forest method for short-term load forecasting

Machine learning methods are main stream algorithms applied in short term load forecasting. However, typical machine learning methods consisting of Artificial Neural Network (ANN) and Support Vector Regression (SVR) have deficiencies hard to overcome, such as easy to be trapped in local optimization (for ANN) or hard to decide kernel parameter and penalty parameter (for SVR). On the other hand, grey relational analysis is an effective method to select proper historical data as training set for training machine learning models. But it is not comprehensive and accurate enough. In this paper, a new two-stage hybrid algorithm aimed to solve these two problems is proposed. Random Forest (RF) method is introduced as the machine learning method, which will not cause overfitting problem and parameters are easy to be tuned. Furthermore, Grey Relational Projection (GRP) is introduced to select similar historical data to train random forest models. The final forecasting results based on real load data prove this new two-stage method performs better than the other two common methods.

Optimizing the boundary of a backup protection in a digital substation using graph theory

With the rapid development of smart grid, digital substation plays an important role to ensure reliability and security of power grid operation. The application of IEC61850 provides the communications standard for total information sharing within a substation. Based on this foundation, this paper analyzes the information from the entire substations primary and secondary equipment and represents the topological structure of digital substation backup protection by means of graph theory. In this way, a graph showing their interconnections with a set of relationship matrices can be produced. Analyzing the relationship matrices can dynamically change the substation differential backup protections boundary in various scenarios including CB opening/closing and CB information loss. The application of graph theory ensures the adaptability and correct operation of substation backup protection. It can also optimize the boundary of the backup protection by reducing the action of breakers.