The power system transients generally include the following three processes: electromagnetic transients, machine-network interactions, and electromechanical transients. These three stages are different from each other and interact with each other. Since the time constants and scopes of the components of the power system and the components are different, traditionally, according to the different purposes of the simulation and the dynamic characteristics of the components, the simulation models and algorithms suitable for a certain stage are selected, and only some of the processes are selected. Simulation. The simulation results using this method are simple and credible within a certain research scope, and have been widely used for a long time. However, with the continuous development of power systems, the scale is expanding, the structure is increasingly complex, and the task of simulation is moving from a simple problem to a complex and diversified one. It is more and more difficult to meet the requirements of simulation research by simply considering the transient characteristics of the system. It is not accurate and reliable to use a single and fixed model to calculate a certain stage of the transient process. Therefore, the whole process of transients is needed. simulation Research.
Establishing an adaptive model of synchronous generators is one of the important parts of transient full-process simulation. Since the generator is a synchronous rotating element, its adaptive model is much more complicated than other stationary elements of the power system. A lot of research has been done on the simulation of the transient full process of the generator at home and abroad. It is very creative to solve the problem of system transient full-process simulation by establishing the interface between the generator and the transient model of the quasi-steady-state model network. However, the following problems still exist: the transient process is still split into different stages, and the generator model is only an integral part of each stage. In fact, only the electromagnetic transient, machine-network interaction and electromechanical transient simulation methods are connected, and no uniform algorithm is formed for the components. Moreover, the transformation of models and algorithms relies on experience and lacks scientific and precise criteria. Such simulation results are not accurate and reliable.
These simulations are still process-oriented simulations rather than object-oriented simulations. The model transformation for process simulation is based on the system process rather than the component object, so that the model cannot be adaptively transformed according to the state of the component. At the same time, there are shortcomings such as readability, modularity, and poor expandability, and it is difficult to achieve unified simulation of the whole process. . With this intelligent simulation model, the speed of simulation can be greatly improved while ensuring accuracy. At the same time, due to the large time constant of the mechanical part, the correction can be predicted when the system frequency does not change much. The system is synchronously solved when the frequency changes such as oscillation and out-of-step are faster. Starting from the above ideas, the object-oriented simulation technology can be used to form a generator adaptive model suitable for transient full-process simulation.
3 Object-Oriented Generator Adaptive Model In this paper, an adaptive transformation criterion based on the characteristics of generator state variables is proposed. The object-oriented technology is used to combine different types of generator models at different stages to form synchronous power generation. Transient adaptive simulation model of the machine.
2 The idea of ​​generator adaptive transient model is to establish a simulation model of the whole process of generator transient, and it is necessary to calculate the whole transient process including electromagnetic transient, machine-network interaction and electromechanical transient. In the simulation process, the component model and the calculation step size are adaptively transformed according to the state of the system and the changing characteristics of the power, so as to ensure the accuracy and simplify the calculation reasonably.
Taking the transient of the no-load three-phase short-circuit fault of the damper generator as an example, the analytical solution of the stator a-phase current can be obtained, which contains the forced component and the free component. The research on the adaptive model of the forced component generator is based on the fourth-order (third-order) model of the rotor which accurately reflects the electromagnetic transient and considers that the rotational speed t is not abrupt. This model is suitable for system-free oscillation (the generator speed t does not change much). ) Initial transient simulation study of faults. Through the analysis of the attenuation component, the simulation model is gradually simplified until a simulation model of the rotor winding dynamics and the stator winding quasi-steady state suitable for electromechanical transient research is formed. In this process, if the system is disturbed again (re-fault or switching operation), it needs to be switched back to the fourth-order (third-order) model of the rotor or the system oscillation model considering the sudden change of the speed t.
In order to truly achieve adaptive simulation of the entire transient process of the power system, each component must be intelligently transformed according to its state. Such traditional process-oriented simulation methods are difficult to perform, so this paper uses object-oriented technology modeling for components (synchronous generators). At the same time, the object-oriented model has the advantages of readability, modularity, and extensibility.
The base class CGeneratoi of the synchronous generator adaptive model provides a unified external interface and public functions, no specific examples. The subclasses derived from the base class CGenerator mainly have the following CGeneratoi subclasses 1 Considering the stator transient and damper windings of the generator electromagnetic transient model. It includes a fourth-order model of the rotor for steam turbines (subclass 1.1) and a third-order model for rotors for turbines (subclass 1.2). Taking the third-order model of the rotor as an example, the attenuation does not occur, while the free component decays with time, and the time constant of the attenuation is small.
Other types of generator transient processes have similar characteristics, so we can consider the electromagnetic process of the stator winding in detail at the initial moment of the fault. When the free component is attenuated enough, the model potential, flux equation and meter are used. The transformation transforms the upper stator part dq0 coordinate equation into abc. The system uses coordinate transformation to transform the generator stator part equation into the synchronous rotation coordinate xy to be connected to the network equation of the generator outlet.
The rotor equation power and moment equations are based on the above equations using the integrated friend model method to obtain the initial transient model of the generator fault. The model accurately considers the electromagnetic transient process.
Transient model of the damper winding. Since the time constant of the stator transient component and the damper winding component of the synchronous generator is small, it is rapidly attenuated. When a component decays to a small enough, the relevant part can be ignored. Subclass 1 and subclass 2.2 formed by corresponding parts are gradually eliminated by subclass 1. Finally, a rotor first-order model without stator transient and damper winding effect can be obtained. The potential and flux linkage equations are described as simulation asymmetry. The system and multiple asymmetric operations, the quasi-steady-state model of the network still uses the abc coordinate system, and the network equation of the generator outlet is called the connected rotor dynamic model. When the system transitions to quasi-steady state, the non-power frequency component is attenuated to a small extent, and the power in the network is basically the fundamental component. At this time, the transient process of the network can be neglected, and the network is described by the algebraic equation of the complex admittance array. At this time, only the rotor dynamics of the generator are taken into account, and the generator is described by a differential equation, so that the machine-network interface is changed accordingly.
Generator stator transients are described by algebraic equations, and other parts (such as rotor field windings, rotor motion equations, etc.) are described by differential equations. The generator equation is differentiated by the previous method and the stator and rotor equations are separated and the initial interface equation of the fault is generated. Generally, the rotor motion equation has a large time constant, and the prediction-correction algorithm can be used when the system frequency does not change much. However, in the case of a complex fault in the system, when the frequency changes rapidly (such as system oscillation), the failure of the frequency change will bring a large error. Therefore, in this paper, a generator synchronization model that solves the rotor motion equation in parallel is used.
Leap variable model. This paper adopts the implicit trapezoidal method modeling, which has good numerical stability, simple calculation and high precision. However, due to the sudden change of the inductor current and the capacitor voltage before and after the fault occurs, the trapezoidal method will cause numerical oscillation. Therefore, the Euler method of spectrum compensation is used to calculate the jump variable: 4. The sub-class 1 corresponding jump variable model (subclass) 6.1) The corresponding jump model corresponding to subclass 5 (subclass 6.2).
In the actual simulation, the history of the transition of the synchronous generator transient adaptive model. The value is obtained from the initial value of the subclass 2, the alEleetroniepublishg transient full-process simulation. The interface criterion for the generator-adaptive transient model into the four subclass transformations and the initial value of the adaptive transformation of the CGenerator class need to be formulated accurately and reasonably. Interface criteria. In the past, the simulation of the variable model was usually based on experience switching at a fixed simulation time, which is very inaccurate. On the one hand, if the model is switched too early and the corresponding transient component is still large, it will cause disturbance and oscillation of the non-original shape; on the other hand, if the switching is too late to avoid the above problem, the calculation amount will be increased. Therefore, based on the variation law of generator state variables, this paper proposes an accurate criterion for adaptive transformation. At the same time, due to the different sub-models, the initial value needs to be re-determined when changing.
41 subclass 6 1 - subclass 1 and subclass 62 - subclass 5 transformation The above two types of transformations are very similar, both from the initial moment of the fault model to the accurate calculation of the electromagnetic transient process of the generator model. The difference is that the former does not consider the change of the rotational speed, and the latter will synchronously solve the dynamic equation of the rotor reflecting the change of the rotational speed. Since the jump variable model is identical to the generator model equivalent circuit that accurately measures the electromagnetic transient process, only the numerical integration method has the same variable, so the historical value can be simply used as the simulation initial value.
4.2 Subclass 1 - Subclass 2 Transformation This transformation process is a process of simplifying the damper winding and transformer potential of the generator model. The transformation criteria are the damper winding current and the rate of change of the flux linkage. Since subclass 2 is a simplification of subclass 1, all variables are a subset of subclass 1, so it can still be transformed by subclass 1 4.3 subclass 2 - subclass 3 due to the network interface with generator model subclass 3 Network transients, so this transformation requires the system to be essentially the fundamental component. The amplitude and phase of the fundamental wave can be obtained by Fourier transform. If the amplitude and phase of the fundamental wave are small in the correlation analysis of the three-phase voltage and current at the interface of the network, the network transient can be ignored. The voltage and current vectors are used as the initial values ​​of the stator windings. For the rotor dynamics and other parts, subclass 3 is the same as subclass 2, so the historical value of subclass 2 can be inherited as the initial value of subclass 3.
4.4 Subclass 3 - Subclass 4 Transform Subclass 3 * Subclass 4 is transformed on the premise that the system undergoes symmetric or asymmetric fault resection, requiring the system to be completely symmetrical. Due to the three-phase symmetry, the vector of the a-phase in the sub-class 3 is the integrated vector in the sub-class 4, that is, the initial value of the other part is the same as the sub-class 3.
4.5 Subclass 2 - Subclass 6 Transformation When the generator is in subclass 2 phase, the system fails again or switches. The generator needs to adopt the jump variable model (subclass 6). Subclass 2 is gradually simplified. When the switch is in any stage, the historical value of the damper winding and the transformer potential has not been simplified. The initial value of subclass 6 is the simplified part of the damper winding current and the initial value of the transformer potential. 0. The initial value of other parts of the generator can also be inherited by historical values.
4.6 Subclass 3 - Subclass 6 and Subclass 4 - Subclass 6 Transformation Since the network is in the subclass 3 phase, the network does not count transients, and is basically the fundamental component. In this way, the instantaneous value of the fundamental wave can be obtained from the vector of the machine-one interface and used as the initial value of the sub-class 6.
From the previous derivation, a is the angle of the a-axis leading the x-axis, and the fundamental voltage vector is U=ReU+j/mU, then the instantaneous value of the a-phase is ua +sinalmU, and the three-phase voltage at the generator outlet can be obtained. The instantaneous value of the current is taken as the initial value of subclass 6. The initial value of the transformer potential and the damper winding current is 0, and other parts can be inherited by historical values. The transformation of subclass 4* subclass 6 can be similarly obtained to obtain a composite vector representing voltage and current.
5 Simulation is a simplified system diagram of 500kV in Central China Power Grid.
A phase-phase grounding short-circuit of A phase occurs in about 370km 500kV of Gezhouba-Phoenix Mountain from about 30% of Gezhouba. 0.09s of the Gezhouba side circuit breaker and the Phoenix Mountain side circuit breaker 01s respectively cut off the faulty line. 0.61s automatic reclosing, after the permanent fault, the above transient process includes multiple faults and switching operations, lasting about 1s, which is a complex process spanning multiple transient phases.
For the Gezhouba outlet transformer system side A phase current waveform and Gezhouba Yigang City generator power angle waveform. Among them, () is the simulation result 13 of the generator transient model using EMTP; (b) is the simulation result using the adaptive model in the text (for comparison, the vector results are converted into instantaneous values) () is Gezhouba Yigang City The power angle difference of the 500kV system diagram motor. Comparing () and (b), it can be seen that the current waveform calculated by the adaptive model reflects the change of the power angle and has higher precision than the EMTP calculation result. At the same time, the computational time complexity is also greatly reduced compared with EMTP. The simulation time ratio of one or more examples is: 0.47:1. This model is especially suitable for electromagnetic transient simulation under system oscillation and transient stability considering the effects of electromagnetic transient processes in complex faults. simulation.
Also non-sinusoidal changes over time. When the power supply voltage is sinusoidal, if the stator leakage impedance voltage drop is neglected, the induced potential of the stator winding is also sinusoidal, and the magnetic density at the edge A of the cutting coil also changes sinusoidally with time. According to the hysteresis loop, the magnetic density can be drawn. The magnetomotive force waveform, for example, is distorted in the waveform. Since the magnetic field rotates in the direction of the phase which is delayed by the leading phase, the Fy lag Fa60* can be calculated from the equation (16) to calculate the current waveform of the phase A winding with time, for example. The resulting waveform is similar to the experimental and simulated waveforms, indicating that the no-load current is asymmetrical in the center line of the half cycle in a positive or negative half cycle, which is caused by hysteresis. The model in this paper can deal with nonlinear problems including hysteresis. . In addition, it can be seen that the no-load current lag potential is large, indicating that the power factor is low at no load; while the load, the stator current increases the active component, making the current waveform close to sinusoidal, and the phase difference between current and back EMF is also reduced, power The factor is increased.
The magnetic density and magnetomotive force at the A and Y sides of the coil and the phase current waveform of the A-phase stator phase 6 conclude that a nonlinear magnetic chain is used to linearize the derivative of the linear flux linkage, and the nonlinear motor system is linearized. Applying the no-load test and numerical analysis method of the motor, the method of obtaining the factor of the nonlinear flux linkage to the derivative of the linear flux linkage is given. The abc basic coordinate system is established to reflect the magnetic saturation, hysteresis and so on. A linear dynamic, new mathematical model of the motor, and simulation results based on this model are consistent with the trend of the motor. When the motor is powered by a sinusoidal power supply, the waveform of the no-load current of the motor is in a positive or negative half cycle due to the influence of hysteresis, and the waveform is asymmetrical to the center line of the positive or negative half cycle. The simulated no-load current waveform is similar to the motor no-load test current waveform, and both have this asymmetry.
When woodworkers talk about CNC cutters, they are most often referring to CNC Routers. To understand what routers can and can`t do, it is important to note that they cut with friction. A cylindrical bit resembling a drill bit with sharp sides spins at high speeds and carves away unwanted material. Unlike a drill, however, a router can cut both in and out and side to side.
cnc machine, cnc router, cnc machinery, cnc router machine, wood cnc router, woodworking machine
JINAN XUANLIN MACHINERY CO., LTD. , https://www.cncfiberlaser.com