What Is Job Safety Analysis?

Safety analysis means that at any time the product or system must consider safety issues, and through analysis to find potential hazards or the possibility of human error. There are several standard methods to analyze this. ^[1]

Security analysis

The Security Analysis function performs predictive accident analysis on a variety of given operating modes (states), and warns of faults that may cause threats to the safe operation of the power grid, such as line overload, voltage overrun and generator power overrun. In order to evaluate the security level of the entire power grid, find the weak links in the system.

For each expected accident, the following three aspects of safety analysis are generally performed:

Calculate the frequency variation of the power system when the large power supply is disconnected or the power system is decommissioned.

On the basis of power flow calculation, check whether the power system components are overloaded and whether the voltage of each node is within the allowable range, that is, check with predetermined safety constraints. If the constraints are not met, an alarm signal and display are sent to the dispatcher.

Perform system stability calculations and check whether the power system can maintain stable operation after the accident.

Anticipated accident sets generally include line breaks, generator and transformer breaks, and three-phase short circuits.

Each expected accident is equivalent to a flow calculation. In order to meet the needs of real-time, fast calculation speed is required, and the calculation accuracy does not have to be too high. Generally, the DC method and PQ decomposition method are used. The power flow calculation of the local area system can provide complete operating information by state estimation, and the external area system calculation can be replaced by external equivalents. External equivalents are usually obtained using the ward node injection method, decoupled ward method, extended ward method, and REI method.

Another important issue of anticipatory accident analysis is how to form a contingency list. The early method required a considerable number of off-line calculations, and then the expected accident items to be executed in each calculation cycle were excluded according to the probability and severity of the accident. Under real-time conditions, it is often difficult to implement all the items specified in the accident table, and only the order can be automatically arranged according to the severity of the accident. That is, under this operating condition, if the specified accident does not cause the system to enter an unsafe state, the analysis and calculation of this calculation cycle will be suspended without having to complete all the specified items.

If the expected accident causes the power system to enter an abnormal state, the safety analysis must take corrective measures to bring the power system into a normal state. This countermeasure is called preventive control.

The usual preventive control always has an economic goal, so that the economic dispatch of the power system under the constraint conditions can be achieved. This is equivalent to solving an optimal power flow problem under constraints. The method was being vigorously studied in the 1980s. A practical algorithm for preventive control is the distribution coefficient method or the conventional DC power flow method. The distribution coefficient method is fast and accurate, but it occupies a large amount of computer storage, while the DC power flow method has poor accuracy.

When the system cannot be found a preventive control solution due to the controllability limitation, the objective function can be to remove the minimum load under the assumption of an emergency situation.

If the safety analysis is performed using the data obtained from the load forecasting, the safety of the system under future conditions can be judged.

The dynamic safety analysis after power system failure generally uses the numerical integration method. The disadvantage of this method is that it has a large calculation workload and cannot give a clear basis for judging the stability of the power system. In order to quickly and stably estimate the stability to meet the requirements of real-time control, the following two methods have been sought.

Lyapunov method: A certain form of Lyapunov function (V function) is established for the stable equilibrium point of the differential equations describing the dynamic process of the power system, and the V function of an unstable equilibrium point in the process of system motion The value is used as an index to measure the size of the stable region near the stable plateau.

Pattern recognition method: Based on off-line simulation calculations of anticipated accidents, a few state variables representing the operating characteristics of the power system are selected to quickly determine the stability of the power system. Using a large number of off-line calculation results for various anticipated accidents, the state space characterizing the operating characteristics of the power system is divided into a stable domain and an unstable domain. In this way, according to the state variable values obtained in real time, it is quickly determined whether the corresponding operation mode is stable in the state space.

The continuous, efficient and secure operation of the network is the basis for the normal operation of the user's business. This requires network managers to be able to grasp key indicators of business application operation at any time, detect abnormalities in time and provide early warning, and implement active operation and maintenance and management. When faults occur, they can quickly and effectively locate problem points, clarify responsibilities, and analyze the cause. Reduce downtime; once the network receives an attack or a security event, it needs to have the means and basis to effectively locate, analyze, and obtain evidence. Kelai Network Backtracking Analysis System is an intelligent, distributed network analysis platform that meets users' ever-increasing needs for network failure, performance, and security analysis.