As fundamental equipment of advanced metering infrastructure (AMI), smart meters collect interval measurements from customer sites and thus provide system wide visibility for power companies. Considering the measurement characteristics of AMI, a series of integrated distribution state estimation methods for medium-voltage and low-voltage distribution network are put forward aiming at online or offline scenarios with 2 kinds of data collection methods respectively. Taking the modified IEEE 13-node test system as an example, the application of the offline scenario with data collected one by one is analyzed and testified. Numerical studies show that the proposed approach can achieve a higher accuracy level for estimation under a shorter cycle, which can provide more effective information for electric power companies.

It is necessary to analyze the impact of various operation conditions on the harmonic of power grid in distributed generation (DG) access programing. In view of large calculation quantity deficiency of Monte-Carlo probabilistic harmonic flow method which need large sample quantity to ensure calculation accuracy, firstly, partial decoupling harmonic flow method is adopted to solve the deterministic harmonic flow, and voltage harmonics distortion rate close to the limit is chosen as kernel samples. Secondly, important sampling method is used to make Latin hypercube sampling have higher probability to focus on the concerned parameter region, thus the size of samples is reduced, then the validity and correctness of the proposed method in IEEE-30 node test system is verified. Finally, the impact of amplitude correlation and harmonic phase randomness of distribution generation on the distortion rate of harmonic voltage in IEEE 33-bus system is computed and analyzed, and it is necessary to consider the amplitude correlation and harmonic phase randomness when optimizing the access of DG.

When the capacity of photovoltaic system is relatively large in DC microgrid, the photovoltaic (PV) system should have the function of bus voltage support. Firstly, a supportive relationship between the output power of PV and the bus voltage of DC microgrid is constituted in this paper, and the real-time photovoltaic output P-V curve is fitted based on sampling values by using the Newton difference method, then the DC voltage control reference value is calculated according to the real-time fitted P-V curve and the relationship between the output power of PV and the bus voltage of DC microgrid. The economy and voltage support capability of photovoltaic generation are considered comprehensively by introducing the output coefficient, and adaptive voltage support is achived by selecting output coefficient through adaptive strategy. Moreover, the maximum power point tracking and constant power control are integrated in this strategy.The principle of this strategy is simple and easy to be realized, it improves the utilization efficiency of PV power generation without narrowing voltage support range for DC bus voltage, and also can enhance the ability for photovoltaic power generation in multiple scenes. Simulation results verify the effectiveness and superiority.

For solving the common existence of feeder overload problem, a load balance technology based on multi-source data fusion is proposed. Firstly, a feeder overload quantitative model is proposed according to the dynamic characteristics of distribution load. Then, distribution load sets are constructed for the three typical application scenerios based on multi-source data of dispatching, operation and maintenance distribution and marketing. Finally, the optimal feeder load balance adjustment strategy is determined through trail calculation and comparison of the overload indexs of multiple distribution reconstruct schemes. Example shows the effectiveness of this technology. Follow this technology, a distribution load management and optimization assistant decision-making system is developed, which has been practically applied to Jiangsu distribution dispatching, operation and maintenance.

As IEC 61850 standard has been maturely applied to intelligent substation, many researchers have tried to apply it to solve the problem of poor inter-operability between distribution terminals. Based on IEC 61850 standard, the modeling problems of transformer terminal unit (TTU) which have rarely been researched are studied in this paper. According to the actual function of TTU, firstly the constructing method of the information model of TTU is expounded, the properties of the model are determined, and then the abstract communication service interface(ACSI) of the model is defined. Secondly taking TTU to instantiate the model, the information model of TTU is established, which expands the applications of IEC 61850 standard to distribution automation systems.

Having a better understanding of the customs of consumers’ electricity consumption and the accurate load forecasting of individual consumer are of great significance to demand response implement and high efficient power system operation. Firstly the load characteristics of large consumers are analyzed, and it is pointed out that the load profiles of large consumers have characteristics such as large in volume and wide in coverage, variant in load features, highly short-term correlative with historic load, notable in fluctuation, and without clear periodic characteristic. According to these characteristics, a wavelet de-nosing and decision tree based method for pattern extraction and load forecasting is proposed. The method mines data of consumer’s historical load and extracts their utilization patterns, then personalizes loads forecasting of consumers based on different utilization patterns. Case studies on 50 typical large consumers in a province of China show that the proposed method is superior to other forecasting methods in accuracy.

In recent years, the development of smart grid in China is gradually promoting, energy storage system (ESS) is one of the key technologies to support the development of smart grid, its economic benefit analysis is very important. This paper improves the life cycle cost (LCC) model of ESD first, then establishes the comprehensive benefit optimization operation strategy and capacity allocation model of ESD based on bilevel programming through analyzing the cost-benefit efficiency which generated by the energy storage device (ESD) of demand side, finally obtains optimal daily operation strategy and capacity of ESD through iteration of the upper and lower model. Simulation analysis of example is carried out by using particle swarm optimization algorithm, the result shows that the outstanding operation strategy and reasonable storage capacity can bring a certain economic benefits, and with the development of the electric market, the installation of energy storage device on the demand side will bring better economic benefits.

In recent years, increasingly tremendous air-conditioning load becomes the main reason why the power consumption of third industry and residen continuously grow, and the seasonal difference between peak and valley load extends. The flexible control of’ central air-conditioning of buildings can be regarded as one of the important peak load regulation accesses, which will improve the power system characteristics and promote it safer and more economic. Firstly, the overall architecture of central air-conditioning of buildings for peak load regulation is proposed, and the typical system reform and deployment are designed. Secondly, flexible control strategies package is constructed from the central air-conditioning itself, and the effect of peak load regulation is quantitatively analyzed. Then, the flexible and optimal decision-making models which satisfy the requirements of comfort degree and peak load regulation are established. At last, the effectiveness of the schemes and model are verified on the project of non-productive air-conditioning control in Jiangsu province.

Considering the general demand of household PV-ESS microgrid, an integrated intelligent microgrid terminal which integrates the functions of bidirectional measurement, supervisory control, energy efficiency management, etc is developed. In order to realize the highly integrated functions of each terminal, the hardware and the algorithm architecture are designed using module method. The hardware aspect design is described by communication between the terminals and the underlying equipment, bidirectional measurement, data storage system, human-computer interaction, etc. Software architecture adopts multi-time scale module classification scheme and system parameters and real time running state are considered to realize the time-scale coordinated control of microgrid equipment. Experimental platform and offline terminal test results show that this integrated terminal can realize the functions of measurement, monitoring and energy management of household PV-ESS microgrid, and it is beneficial to popularize household PV-ESS microgrid application.

Lead-acid battery manufacturers belong to the high power energy consumption industry, and they have great energy-saving potential by improving formation process. Considering the peak-valley electricity pricing mechanism, and aiming at minimizing the total power cost of battery packs during a formation cycle, a mathematic optimal dispatching model of lead-acid battery formation is established based on the lead-acid battery formation process, with the premise that there is no feedback to superior power grid and no additional large-scale energy storage devices installed. The conversion efficiency of charging-discharging device is considered together with the constraints of load power of charging-discharging devices, no feedback to power grid, switching time interval of battery packs, switching frequency during a formation dispatch cycle. Based on the actual operation data of a lead-acid battery factory, the particle swarm optimization algorithm is used to solve the optimal model. Results show that the total power consumption cost can be reduced by 5% after optimization. The research in this paper can provide reference for lead-acid battery industry in energy-saving reconstruction and efficiency promotion.

Non-intrusive load monitoring (NILM) is a novel detailed power load consumption monitoring technology. Research of electrical appliance load characteristics is the foundation of realizing NILM. Along with the increase of electronic and power electronic appliances equipment, harmonic characteristics have gradually become the focus of NILM. For a better application of harmonic characteristics to NILM, firstly, two indices for measuring the repeatability and discrimination of load characteristics are proposed. Based on the research on harmonic generation mechanism of electronic appliances, six kinds of harmonic characteristic indices are summerized, and their repeatability and discrimination are analyzed in detail according to the proposed two indices. Sample characteristics of different kinds and different brands of appliances from five family scenarios and two business scenarios are summerized, and the measured dataset is established. Analysis on the measured dataset shows that the 3rd and 5th harmonic amplitude value and total harmonic distortion have better repeatability and discrimination according to the proposed evaluation indices of harmonic characteristics. This conclusion corresponds to the accuracy results by different classifiers established on different harmonic characteristics, therefore it can not only guide the effective application of harmonic characteristics to NILM, but also prove the effectiveness of the two proposed load characteristics evaluation indices.

Because the actual distribution network reconstruction project funds are often limited, the cost-benefit model and the decision-making method are crucial for distribution network energy saving reconstruction project. From the perspective of life cycle cost(LCC), firstly the research life cycle is determined for the energy saving reconstruction of distribution networks with multi-devices. Then, a new life cycle cost-benefit model for energy-saving reconstruction of distribution network is developed, in which the modification schemes include distribution transformers replacement, lines replacement and reactive power compensation. In the operation loss cost and maintenance cost area, the operation cost model considering the influence of load season characteristics and the maintenance cost segmental model of transformers are proposed. Finally, aiming at the highest energy saving profit per LCC, a decision-making method is developed while considering financial and technical constraints as well. The model and method are applied to a real distribution network reconstruction, and the results prove that the model and method are effective.