Simulation system of power conversion equipment and simulation method thereof

Abstract

The invention provides a simulation system of power conversion equipment and a simulation method thereof, and belongs to the technical field of power electronics. An upper computer in the system is used for issuing a circuit model of the power conversion equipment to a programmable logic array; and the programmable logic array is used for updating the current values of the current sources in the plurality of switch branches at the current simulation moment according to the branch voltage values and the branch current values of the plurality of switch branches at the previous simulation moment in the simulation process of the circuit model. Wherein the current value I (t) of the current source in each switch branch at the current simulation moment meets the condition that I (t) = alpha * G * v (t-t) + beta * i (t-t) + gamma; alpha is a voltage coefficient, G is an admittance value of the switch branch, v (t-t) is a branch voltage value at the previous simulation moment, beta is a current coefficient, i (t-t) is a branch current value at the previous simulation moment, gamma is a current compensation value, and t is a simulation step length. And the simulation precision of the simulation system is improved.

Description

Technical Field

[0001] The present application relates to the field of power electronics technology, and in particular to a simulation system and a simulation method for a power conversion device. Background Art

[0002] Power electronic equipment products, from R&D, testing, certification, to actual project deployment, rely on real-time simulation technology. Power electronic equipment refers to devices that utilize power electronics technology to convert and control electrical energy. Examples include inverters that convert direct current to alternating current, or frequency converters that convert fixed-frequency AC to adjustable-frequency AC. Because power conversion equipment primarily relies on the switching characteristics of power electronic devices to achieve power conversion and control, and power electronic devices with switching characteristics are called switching elements, simulating switching elements is crucial for simulating power conversion equipment.

[0003] In related technologies, the constant admittance method is used to model switching elements. When the switching element is in the on state, it is equivalent to an inductor, and when the switching element is in the off state, it is equivalent to a capacitor. The equivalent admittance of the inductor and capacitor is equal to avoid frequent changes in the switch element's admittance caused by high-frequency switching.

[0004] However, each time the switching element switches state, the inductor or capacitor must be recharged due to the different voltage and current characteristics of the inductor or capacitor, resulting in a loss of energy within the switching element during the switching process. Consequently, virtual power loss is inevitable, leading to a slow transient response and transient errors, which affect simulation accuracy. Summary of the Invention

[0005] The present application provides a simulation system and a simulation method for a power conversion device, which can solve the problem of low simulation accuracy in related technologies.

[0006] In a first aspect, a simulation system for a power conversion device is provided, the system comprising: a programmable logic array (PLA) and a host computer; the host computer is used to send a circuit model of the power conversion device to the programmable logic array, the circuit model comprising multiple switch branches, each switch branch being used to be equivalent to a switch element, and each switch branch comprising a current source and admittance connected in parallel; the programmable logic array is used to, during a simulation of the circuit model, update the current values of the current sources in the multiple switch branches at a current simulation moment according to the branch voltage values and branch current values of the multiple switch branches at a previous simulation moment; and update the branch current values and branch voltage values of the multiple switch branches at a current simulation moment according to the admittance values of the multiple switch branches and the current value of the current source at the current simulation moment. Among them, the current value I(t) of the current source in each switching branch at the current simulation moment satisfies: I(t) = α*G*v(t-△t)+β*i(t-△t)+γ; α is the voltage coefficient, G is the admittance value of the switching branch, v(t-△t) is the branch voltage value at the previous simulation moment, β is the current coefficient, i(t-△t) is the branch current value at the previous simulation moment, γ is the current compensation value, and △t is the simulation step size.

[0007] In the solution provided by this application, the switching elements in the circuit model are equivalent to parallel current sources and admittances. At each simulation moment, the current value of the current source is updated according to the branch voltage value and branch current value at the previous simulation moment, thereby avoiding virtual power loss during the switch state switching process. Moreover, in the process of updating the current value of the current source, not only the rate at which the switch branch converges to a steady state is increased by the voltage coefficient and the current coefficient, but also the initial state after the switch state is switched is made closer to the steady state value by the current compensation value. This effectively shortens the time for the switch branch to converge to a steady state after the switch state is switched, avoids transient errors caused by slow transient response, and improves the simulation accuracy of the simulation system.

[0008] In one possible implementation, the voltage coefficient, current coefficient, and current compensation value corresponding to the switch branch in the on state are different from those corresponding to the switch branch in the off state. This allows the current source to be updated in a manner that matches the switch state of the switch branch, improving the simulation accuracy of the switch branch in different switching states.

[0009] In one possible implementation, the current coefficient of the switch branch in the on-state and the voltage coefficient in the off-state are values obtained by solving a discrete system model based on the final value theorem and steady-state characteristics. The discrete system model is a complex frequency domain relationship between branch voltage and branch current generated based on a relationship for the current value I(t). The steady-state characteristics include a steady-state voltage of 0 in the on-state and a steady-state current of 0 in the off-state. Thus, by converting the circuit model into a discrete system model for a discrete-time system, a solution is obtained for the discrete-time system that satisfies the steady-state characteristics of the power conversion device based on the final value theorem, so that the response process simulated based on the current coefficient in the on-state and the voltage coefficient in the off-state meets the requirements of the steady-state characteristics, thereby improving the simulation accuracy of the simulation system.

[0010] In one possible implementation, the voltage coefficient of the switch branch in the on-state and the current coefficient in the off-state are coefficients within a range where the spectral radius is less than a threshold value. The spectral radius is the spectral radius of the state matrix of the current source iterated based on the relationship between the current value I(t). Thus, by analyzing the iterative state matrix of the current source in the circuit model, the voltage coefficient in the on-state and the current coefficient in the off-state can be selected within a range where the spectral radius is less than the threshold value. Since a smaller spectral radius results in a faster convergence rate, the selected voltage coefficient in the on-state and current coefficient in the off-state correspond to the fastest convergence rate.

[0011] In one possible implementation, the current compensation value of the switch branch in the on state is less than 0 and greater than the negative value of the steady-state current value of the switch branch in the on state; the current compensation value of the switch branch in the off state is greater than 0 and less than the product of the steady-state voltage value of the switch branch in the off state and the admittance value of the switch branch.

[0012] Since the current value of the steady-state current source when the switch branch is in the on state is the negative of the steady-state current value of the switch branch, a current compensation value close to the steady-state current source is selected to compensate the current source. Similarly, since the current value of the steady-state current source when the switch branch is in the off state is the product of the steady-state voltage value and the admittance value of the switch branch, a current compensation value close to the steady-state current source is selected to compensate the current source. Thus, the current compensation value determined in the above manner makes the current value of the current source of the switch branch when switching between on and off states closer to the current value of the steady-state current source, thereby shortening the convergence time of the transient response.

[0013] In one possible implementation, a programmable logic array is configured to: determine the input current value of each node of a circuit model based on the current value of a current source of multiple switch branches at a current simulation time, where a node refers to a connection point between two adjacent switch branches in the circuit model; determine the node voltage value of each node based on the admittance values of the multiple switch branches and the input current value of each node; and update the branch current values and branch voltage values of the multiple switch branches at the current simulation time based on the node voltage values of each node. This makes the determined branch current values and branch voltage values more accurate.

[0014] In a possible implementation, the admittance value G of the admittance of the switching branch satisfies:

[0015]

[0016] Where R is the damping resistance of the switch branch in the off state, C is the capacitance of the switch branch in the off state, and L is the inductance of the switch branch in the on state. Therefore, by retaining an additional damping resistor in the capacitance branch, virtual oscillations caused by unstable capacitance or inductance are avoided, making the determined admittance value more accurate.

[0017] In a second aspect, a simulation method for a power conversion device is provided, the method comprising: simulating a circuit model of the power conversion device, the circuit model comprising a plurality of switch branches, each switch branch being used to be equivalent to a switch element, and each switch branch comprising a current source and admittance connected in parallel; in the process of simulating the circuit model, updating the current values of the current sources in the plurality of switch branches at the current simulation moment according to the branch voltage values and branch current values of the plurality of switch branches at the previous simulation moment; and updating the branch current values and branch voltage values of the plurality of switch branches at the current simulation moment according to the admittance values of the plurality of switch branches and the current value of the current source at the current simulation moment.

[0018] Among them, the current value I(t) of the current source in each switching branch at the current simulation moment satisfies: I(t) = α*G*v(t-△t)+β*i(t-△t)+γ; α is the voltage coefficient, G is the admittance value of the switching branch, v(t-△t) is the branch voltage value at the previous simulation moment, β is the current coefficient, i(t-△t) is the branch current value at the previous simulation moment, γ is the current compensation value, and △t is the simulation step size.

[0019] In a possible implementation, the voltage coefficient, current coefficient, and current compensation value corresponding to the switch branch in the on state are different from the voltage coefficient, current coefficient, and current compensation value corresponding to the switch branch in the off state.

[0020] In one possible implementation, the current coefficient of the switch branch in the on state and the voltage coefficient in the off state are values obtained by solving a discrete system model based on the final value theorem and steady-state characteristics. The discrete system model is a relationship between branch voltage and branch current in the complex frequency domain generated based on a relationship of the current value I(t). The steady-state characteristics include a steady-state voltage of 0 in the on state and a steady-state current of 0 in the off state.

[0021] In one possible implementation, the voltage coefficient of the switch branch in the on state and the current coefficient in the off state are coefficients within a range where the spectrum radius is less than a threshold value, and the spectrum radius is the spectrum radius of the state matrix of the current source based on the iterative relationship of the current value I(t).

[0022] In one possible implementation, the current compensation value of the switch branch in the on state is less than 0 and greater than the negative value of the steady-state current value of the switch branch in the on state; the current compensation value of the switch branch in the off state is greater than 0 and less than the product of the steady-state voltage value of the switch branch in the off state and the admittance value of the switch branch.

[0023] In one possible implementation, the branch current values and branch voltage values of the multiple switch branches at the current simulation moment are updated based on the admittance values of the multiple switch branches and the current value of the current source at the current simulation moment, including: determining the input current value of each node of the circuit model based on the current value of the current source of the multiple switch branches at the current simulation moment, where a node refers to a connection point between two adjacent switch branches in the circuit model; determining the node voltage value of each node based on the admittance values of the multiple switch branches and the input current value of each node; and updating the branch current values and branch voltage values of the multiple switch branches at the current simulation moment based on the node voltage value of each node.

[0024] In a possible implementation, the admittance value G of the admittance of the switching branch satisfies:

[0025]

[0026] Wherein, R is the damping resistance value of the switch branch in the off state, C is the capacitance value of the switch branch in the off state, and L is the inductance value of the switch branch in the on state.

[0027] The technical effects that can be achieved by any possible implementation of the second aspect can be referred to the technical effects that can be achieved by any possible implementation of the first aspect, and will not be repeated here. These and other aspects of the present application will be more concise and easy to understand in the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1This is a schematic diagram of the structure of a simulation system for a power conversion device provided in an embodiment of the present application;

[0029] Figure 2 This is a schematic diagram of a circuit model provided in an embodiment of the present application;

[0030] Figure 3 This is a schematic diagram of an equivalent structure of a switch element provided in an embodiment of the present application;

[0031] Figure 4 1 is a schematic structural diagram of a step-down chopper circuit provided in an embodiment of the present application;

[0032] Figure 5 This is a spectral radius contour map provided in an embodiment of the present application;

[0033] Figure 6 1 is a comparative schematic diagram of a state transition trajectory of a historical current source provided in an embodiment of the present application;

[0034] Figure 7 is a structural diagram of another circuit model provided in an embodiment of the present application;

[0035] Figure 8 This is a schematic diagram of a simulation process of a programmable logic array for a circuit model provided by an embodiment of the present application;

[0036] Figure 9 This is a flow chart of a simulation method for a power conversion device provided in an embodiment of the present application;

[0037] Figure 10 This is a schematic diagram of a simulation process of a power conversion device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0038] The following describes in detail the simulation system and simulation method of the power conversion device provided by the embodiment of the present application in conjunction with the accompanying drawings. First, the key terms involved in the embodiment of the present application are introduced.

[0039] Real-time simulation: Real-time simulation is a type of simulation mode, corresponding to non-real-time offline simulation. It is a simulation mode in which the running time of the simulation model is strictly synchronized with the real physical world.

[0040] Hardware-in-the-loop (HIL) simulation: Hardware-in-the-loop (HIL) simulation is a subset of real-time simulation. A real-time simulation system can be constructed as a HIL simulation system. A HIL simulation system connects a real-time simulation model to actual hardware devices via input / output (IO) interface boards, forming a HIL and exchanging data in real time.

[0041] Power conversion equipment: A power conversion device is a device used to convert and control electrical energy, capable of converting between different power parameters to meet the specific needs of various electrical devices and systems. Power conversion equipment may include, but is not limited to, transformers, frequency converters, rectifiers, or inverters. An example of this is photovoltaic inverters used in new energy power electronics.

[0042] With the increasing number of power electronic devices connected to the power grid and the rapid development of semi-physical hardware-in-the-loop simulation, the simulation accuracy and speed of real-time simulators are difficult to meet user needs. Due to the high-frequency switching characteristics of the switching elements in power conversion equipment, in order to complete the real-time simulation of power conversion equipment, the real-time simulation equipment is required to have a small-step real-time simulation capability. The small-step real-time simulation capability refers to the ability to simulate and calculate the system in real time with a very small simulation step size during the simulation process. Usually, the simulation step size of power electronic systems is in the microsecond (μs) or below. To achieve the small-step real-time simulation capability, the real-time simulation equipment will use a field-programmable gate array (FPGA) as the main simulation computing hardware.

[0043] In this case, the high-frequency switching characteristics of power conversion devices lead to parameter switching during real-time simulation. This presents a major challenge for devices using FPGAs as the primary computing hardware for real-time simulation and is a core issue that FPGA real-time simulation must address. In one possible implementation, a constant admittance modeling approach can be used for the switching elements in power conversion devices to avoid the parameter switching issue.

[0044] The constant admittance modeling method involves the following: when the switch is on, the switch is equivalent to an inductor, L; when the switch is off, the switch is equivalent to a capacitor, C. The equivalent L / C of the switch is discretized through numerical integration, and the switch is then equivalent to a parallel connection of an admittance and a current source. The values of the inductor and capacitor are adjusted to ensure that the admittance values of the equivalent admittance remain constant during on- and off-state operation, thus avoiding parameter switching during real-time FPGA simulation. However, each time the switch state changes, the inductor or capacitor needs to be recharged, which not only causes overpulsing in the switching waveform but also generates a large amount of virtual power loss, leading to virtual losses and transient oscillations at high switching frequencies.

[0045] See also Figure 1 , Figure 1 This is a schematic diagram of the structure of a simulation system for a power conversion device provided in an embodiment of the present application. Figure 1As shown, the simulation system may include: a host computer and a programmable logic array. Among them, the host computer stores models corresponding to various components, for example, models corresponding to resistors, capacitors, inductors and switching elements can be stored. Users can call the model in the host computer according to needs to obtain the circuit model of the power conversion device, and send the circuit model to the programmable logic array. For example, the host computer is electromagnetic transient simulation software. The programmable logic array (PLA) is used to simulate the circuit model, and the simulation in the embodiment of the present application may refer to electromagnetic transient simulation. For example, PLA can be a complex programmable logical array (CPLA), FPGA, generic array logic (GAL) or any combination thereof.

[0046] Optionally, the circuit model includes multiple elements and the connection relationship between the elements. Each element can be considered as a branch. For example, a switch element can be considered as a switch branch in the circuit model, and a capacitor element can be considered as a capacitor branch in the circuit model. The connection point between two adjacent branches in the circuit model is a node. Usually, the circuit model includes a reference node. The potential values of other nodes are relative to the potential value of the reference node. For example, the potential value of the reference node can be 0. For example, see Figure 2 The circuit model shown includes reference node 0, node 1, node 2 and node 3. The power branch of the voltage source Us is between the reference node 0 and node 1, the capacitance branch of the capacitor C is between the reference node 0 and node 2, the resistance branch of the resistor R is between the reference node 0 and node 3, the inductance branch of the inductor L1 is between the node 1 and node 2, and the inductance branch of the inductor L2 is between the node 2 and node 3.

[0047] In one possible implementation, a host computer is used to number circuit nodes based on the circuit topology and component parameter information of the power conversion device to form a circuit netlist for the power conversion device. In this way, the circuit topology is converted into a digital model that can be processed by a computer. A circuit netlist is a file that describes the circuit structure and component parameters in text form and contains information about each component in the circuit, where the information for each component includes the node connection relationship of the component and the parameter value of the component. For example, a circuit netlist may include "R1 1 2 100 ohms", which means that resistor R1 is connected between node 1 and node 2 and has a resistance of 100 ohms (Ω).

[0048] The circuit model of the power conversion device includes at least one switch branch, each switch branch is used to be equivalent to a switch element, and each switch branch includes a current source and an admittance in parallel. Among them, the switch element refers to an element with switching characteristics, including but not limited to a power diode, an insulated gate bipolar transistor (IGBT) or a metal-oxide-semiconductor field effect transistor (MOSFET), etc., and the embodiment of the present application does not limit this. Optionally, the circuit model of the power conversion device also includes any other branches such as a capacitor branch, a resistor branch or an inductor branch, and the embodiment of the present application does not limit this.

[0049] In one possible embodiment, the host computer is used to discretize the mathematical model of each element connected between each node according to the circuit netlist of the power conversion device, and abstract it into an equivalent model suitable for solving by the simulation algorithm. Optionally, the host computer is used to use the trapezoidal integration method, the forward and backward Euler method or the backward Euler method to perform numerical integration and discretization on each switching element in the circuit model, and convert it into a unified current source parallel admittance structure, and the structure of the current source parallel admittance is the model corresponding to the switching element. In the case where the circuit model also includes other elements besides the switching element, the host computer is also used to use the trapezoidal integration method, the forward and backward Euler method or the backward Euler method to perform numerical integration and discretization on other elements in the circuit model, and convert them into a unified current source parallel admittance structure.

[0050] In the embodiment of the present application, the switch element in the on state can be equivalent to an inductor, and the switch branch in the on state can be equivalent to an inductor branch; the switch element in the off state can be equivalent to a capacitor, and the switch branch in the off state can be equivalent to a capacitor branch. Taking the capacitor branch as an example, the current-voltage relationship of the capacitor is i C =C*dv C / dt,i C is the branch current value of the capacitor branch, C is the capacitance value, v C is the branch voltage value of the capacitor branch. According to the backward Euler method dv C / dt≈(v C (t)-v C (t-△t)) / △t, then i C (t)=C*(v C (t)-v C (t-△t)) / △t, △t represents the discrete step length, which is consistent with the simulation step length in the simulation system, and t represents the current moment. The following formula (1) can be obtained by sorting.

[0051]

[0052] Therefore, the capacitor branch can be equivalent to a current source With an admittance Similarly, the inductor branch is discretized by mathematical integration using the backward Euler method, and the following formula (2) can be obtained. Therefore, the inductor branch can be equivalent to a current source I L (t)=-i L (t-△t) and an admittance The parallel structure of L is the branch current value of the inductor branch, L is the inductance value, v L is the branch voltage value of the inductor branch.

[0053]

[0054] Through mathematical integral discretization, the switch branches in different switching states can be equivalent to the structure of current source parallel admittance. By setting the capacitance and inductance values, the admittance value G in the open state is L and the admittance value G in the off state C Same, that is, G=G L =△t / L=G C =C / △t to avoid changes in admittance caused by switching states during simulation. However, the inductance value in the on state and the capacitance value in the off state do not always make the admittance value the same. Moreover, the capacitance and inductance values also need to simulate the transient behavior of the switching elements, which is also the main cause of virtual oscillation.

[0055] The embodiment of the present application adjusts the value of the admittance value of the switch branch, and retains an additional damping resistor on at least one of the capacitive branch in the off state or the inductive branch in the on state. The damping resistor is a resistor element used to limit the rate of change of current in the circuit and suppress oscillation or overvoltage phenomena that may occur in the circuit. For example, the admittance value G of the admittance of the switch branch satisfies the following formula (3). Wherein, R is the damping resistor value of the switch branch in the off state, C is the capacitance value of the switch branch in the off state, L is the inductance value of the switch branch in the on state, and △t is the simulation step size. Therefore, by retaining the additional damping resistor, virtual oscillation caused by unstable capacitance value or inductance value is avoided, so that the determined admittance value is more accurate.

[0056]

[0057] For example, the switch branch can be equivalent to an inductor L in which the switch element is in the on state. Figure 3 In (a), the left side represents the inductor L, and the right side represents the model equivalent to the inductor L, that is, the current source I LThe structure of parallel admittance G. Similarly, the switch branch can be equivalent to a capacitor C in the off state of the switch element, Figure 3 In (b), the left side represents the capacitor C, and the right side represents the model equivalent to the capacitor C, that is, the current source I C The structure of the parallel admittance G. Alternatively, the switch branch can also make the switch element in the on state equivalent to an additional element such as an inductor connected in parallel or a resistor connected in series, and the switch branch can also make the switch element in the off state equivalent to an additional element such as a capacitor connected in parallel or a resistor connected in series.

[0058] After the mathematical model of each component in the circuit model is discretized, each component can be equivalent to a current source parallel admittance structure. Since the switch branch can be equivalent to a capacitor branch or an inductor branch in different switching states, and according to the above, the current source I equivalent to the inductor branch L (t)=-i L (t-△t), the equivalent current source I of the capacitor branch C (t) = G*v C (t-△t), the current value of the current source of the switch branch satisfies the following formula (4). That is, when the switch branch is in the on state (s=1), the current value of the current source at the current simulation time t is determined based on the branch current value i(t-△t) at the previous simulation time (t-△t); when the switch branch is in the off state (s=0), the current value of the current source at the current simulation time t is determined based on the branch voltage value v(t-△t) at the previous simulation time (t-△t).

[0059]

[0060] Since the current value of the equivalent current source is related to the branch voltage and branch current at the historical simulation moment, the equivalent current source can be called a historical current source. In the embodiment of the present application, the update formula of the historical current source of the switch branch can be expressed as a unified form shown in formula (5), so that when facing the switching of different switch states, the simulation system updates the historical current source in the same formula form, avoiding the parameter switching of hardware simulation. Wherein, α is the voltage coefficient, G is the admittance value of the switch branch, u(t-△t) is the branch voltage value at the previous simulation moment, β is the current coefficient, i(t-△t) is the branch current value at the previous simulation moment, γ is the current compensation value, and △t is the simulation step size.

[0061] I(t)=α*G*v(t-△t)+β*i(t-△t)+γ Formula (5)

[0062] In an embodiment of the present application, the voltage coefficient and the current coefficient can be determined based on the steady-state characteristics and transient characteristics of the switching element, so as to make the model corresponding to the switching element meet the steady-state characteristics and improve the convergence speed of the transient response; the compensation power value can be determined based on the steady-state value of the switching element, so as to make the initial state of the model corresponding to the switching element closer to the steady-state value after the state is switched. Thus, while improving the convergence speed by the voltage coefficient and the current coefficient, the initial state is also changed by the compensation power value, so as to shorten the time for the switching element to converge to the steady state from different aspects, avoid the limited improvement of the transient response caused by only the voltage coefficient and the current coefficient, and avoid the insufficient compensation caused by only the compensation power value. Since one switch branch corresponds to one switch element, the switching state of the switch branch is also the switching state of the corresponding switch element.

[0063] Among them, the steady-state characteristics include the steady-state voltage of 0 in the on state and the steady-state current of 0 in the off state. The transient characteristics include the trend of the branch voltage or branch current converging to the steady-state value after the switch state is switched. The steady-state value refers to the branch voltage value or branch current value after the switching element reaches the steady state. In the on state (on): the steady-state voltage u on (+∞)=0, steady-state current i on (+∞) is determined by the external characteristics of the switching element; in the off state (off): steady-state voltage u off (+∞) is determined by the external characteristics of the switching element, the steady-state current i off (+∞) = 0. For example, when the switching element is an IGBT, the steady-state current when the IGBT is turned on is affected by factors such as the collector-emitter voltage, the gate voltage, and the physical characteristics of the chip.

[0064] Furthermore, the update formula of the historical current source under different switching states can be shown as formula (6). That is, the voltage coefficient α corresponding to the switch branch in the on state (s=1) is on , current coefficient β on , current compensation value γ on The voltage coefficient α corresponding to the off state (s=0) off , current coefficient β off , current compensation value γ off Therefore, the updating mode of the current source is matched with the switching state of the switch branch, thereby improving the simulation accuracy of the switch branch in different switching states.

[0065]

[0066] Next, we will analyze α on , β on , γ on , α off , βoff , γ off The selection of the value of is given as an example.

[0067] (1) Determine the voltage coefficient α in the off state based on the steady-state characteristics off and the current coefficient β in the on-state on .

[0068] In one possible implementation, the current coefficient of the switch branch in the on-state and the voltage coefficient in the off-state are values obtained by solving a discrete system model based on the final value theorem and steady-state characteristics. The discrete system model is a complex frequency domain relationship between the branch voltage and the branch current generated based on a relationship for the current value I(t) of the current source. The relationship for the current value I(t) of the current source is the aforementioned formula (5).

[0069] For example, the above-mentioned mathematical integral discretization of the switch branch using the backward Euler method is used as an example. and admittance Substitute the above formula (1) and let the current source I L (t)=-i L (t-△t) and admittance Substituting the above formula (2), we can get the calculation formula of the branch variable of the switch branch as shown in the following formula (7). Among them, the branch variable of the switch branch includes the branch voltage value v in the on state C (t) and the branch current value i in the off state L (t).

[0070]

[0071] Furthermore, the change process of the branch current value of the switch branch in the on state and the off state can be uniformly expressed as the relationship shown in formula (8). Among them, i(t) represents the branch current value of the switch branch at the current simulation time, G represents the admittance value of the switch branch, v(t) represents the branch voltage value of the switch branch at the current simulation time, and I(t) represents the current value of the current source of the switch branch at the current simulation time.

[0072] i(y)=G*v(t)-I(t) Formula (8)

[0073] Combining the above formulas (6), (7) and (8), the discrete system model of the switch branch is updated to the following formula (9).

[0074]

[0075] Among them, the branch variables of the switch branch need to satisfy the steady-state characteristics, that is, the steady-state voltage in the on state is 0, and the steady-state current in the off state is 0. For the discrete system model shown in formula (9), the steady-state voltage in the on state and the steady-state current in the off state are calculated through Z-transform and the final value theorem. The final value theorem is used to solve the limit value of the discrete sequence as t→+∞. If the Z-transform of the discrete system v(t) is U on (z), and all the poles of U on (z) are inside the unit circle of the Z-plane, then the final value v(+∞) of the discrete system v(t) = lim z→1 U on (z)(z - 1). Thus, the value of the discrete sequence at steady state can be directly obtained through the final value theorem, without first finding the analytical expression of the discrete sequence and then taking the limit.

[0076] Performing Z-transform on the discrete system model shown in formula (9), the relational expressions in the complex frequency domain shown in the following formula (10) and formula (11) are obtained. Formula (10) is the relational expression between the branch voltage and the branch current in the complex frequency domain in the on state, and formula (11) is the relational expression between the branch voltage and the branch current in the complex frequency domain in the off state. Among them, U on and U off are the branch voltage values in the on state and the off state respectively, I on and I off are the branch current values in the on state and the off state respectively, and z -1 is a unit delay link.

[0077]

[0078] Furthermore, through the final value theorem, it can be obtained that the steady-state voltage value U on (+∞) is as shown in the following formula (12), and the steady-state current value I off (+∞) is as shown in the following formula (13). To make formula (12) and formula (13) satisfy the steady-state characteristics of U on (+∞) = 0 and I off (+∞) = 0, the necessary and sufficient condition that the numerator is 0 but the denominator cannot be 0 needs to be satisfied. Then α on ≠ 1, β on = -1; α o ff = 1, β o ff ≠ -1. Thus, it can be determined that α o ff = 1 and β on = -1.

[0079]

[0080] (2) Determine the voltage coefficient α in the on state based on the transient characteristics on and the current coefficient β in the off state off .

[0081] In a possible implementation, the voltage coefficient in the on state and the current coefficient in the off state of the switch branch are coefficients within a range where the spectral radius is less than a threshold. The spectral radius is the spectral radius of the state matrix iterated based on the relationship of the current source with respect to the current value I(t). In the embodiments of this application, based on the relationship of the current value I(t) of the historical current source shown in formula (5), with the historical current source of the switch branch as the state variable, the state equation for the iteration of the historical current source can be listed according to the topological structure of the circuit model and Kirchhoff's laws. Among them, Kirchhoff's laws include Kirchhoff's current law (KCL) and Kirchhoff's voltage law (KVL). KCL includes that the sum of the currents entering a node is equal to the sum of the currents leaving the node, and KVL includes that the algebraic sum of the voltages across all elements along a closed loop is equal to zero.

[0082] Since α o ff = 1 and β on = -1 have been determined, the state matrix of this state equation includes α on and β off . Furthermore, the spectral radius of the state matrix under different values of α on and β off can be obtained, and α on and β off within a range where the spectral radius is less than the threshold are selected. Among them, the threshold can be set according to experience or flexibly adjusted according to the application scenario. For example, the threshold can be 1. The spectral radius refers to the maximum value of the modulus of the eigenvalues of the state matrix, and the size of the spectral radius reflects the convergence speed of the system. The smaller the spectral radius, the faster the system converges, and the larger the spectral radius, the slower the system converges. Optionally, by plotting the contour lines of the spectral radius, the influence of α on and β off on the convergence speed can be visually seen. The contour lines of the spectral radius refer to the curves formed by connecting points with the same spectral radius in the state matrix.

[0083] Exemplarily, taking the buck circuit shown in Figure 4 as the circuit model, the circuit model includes two switching elements, namely a semiconductor element and a freewheeling diode, and the circuit model also includes elements such as a voltage source, a capacitor, a resistor, and an inductor. Since the current in the inductor hardly changes at the moment of switch switching, the inductor can be considered as a current source. Using the response matching model for the switching element shown in Figure 4 , with the historical current source of the switching element as the state variable, the state equation for the iteration of the historical current source sorted out according to the connection relationship of the circuit topology can be as shown in the following formula (14).

[0084]

[0085] Among them, I1 and I2 are the current values of the historical current source of the semiconductor element and the freewheeling diode, v1 and v2 are the branch voltage values of the semiconductor element and the freewheeling diode, i1 and i2 are the branch current values of the semiconductor element and the freewheeling diode, and A is the state vector The state matrix, also called the state transfer matrix, B is the input vector The input matrix.

[0086] In the case where the semiconductor element and the freewheeling diode are in different combinations of switching states, different state matrices can be obtained according to Kirchhoff's law. When the semiconductor element is in the state and the freewheeling diode is in the off state, the state matrix A can be shown as follows (15). Alternatively, when the semiconductor element is in the off state and the freewheeling diode is in the on state, the state matrix A can be shown as follows (16). Among them, different state matrices under different combinations of switching states are equivalent, for example, Figure 5 is a spectral radius contour plot based on formula (15) or the state matrix A shown in formula (15).

[0087]

[0088] Theoretically, the smaller the spectral radius, the faster the system converges to a steady state. If there is a point minimum value where the spectral radius is 0, a set of coefficients with the smallest spectral radius is selected. Figure 5 The contour map shown includes a contour line with a spectral radius of 0, so the coefficient corresponding to the case of a spectral radius of 0 is selected. or, However, not all circuit connection topologies and parameters can arbitrarily configure the spectrum radius to the minimum value of 0. There is a possibility that the minimum spectrum radius of some circuit models will not be very close to 0 or even close to 1. In the embodiment of the present application, it is sufficient to flexibly select values within the range of the small threshold of the spectrum radius. For example, Figure 5 The shaded area in the figure is the area surrounded by the contour line with a spectrum radius of 1. You can flexibly select α in this shaded area. on and β off The value of .

[0089] (3) Determine the current compensation value γ in the on-state based on the steady-state value on and the current compensation value γ in the off state off .

[0090] According to the above formula (8), the relationship between the current value I(t) of the historical current source at the current simulation moment and the branch current value i(t) at the current simulation moment is I(t)=G*v(t)-i(t).

[0091] In the on state, the steady-state voltage of the switch branch is 0, that is, v s =0, then the steady-state value of the historical current source in the on-state is I ons =-i s ,i s is the steady-state current value in the on-state. Therefore, the expression of the historical current source I(t) shown in formula (6) is I(t) = α when it reaches the on-state steady state. on *G*v(t-△t)+β on *i(t-△t)+γ on =-i s When the switch state switches from off to on, since i(t-△t)=0, the initial value of the historical current source after switching is I(t)=α on *G*v(t-△t)+γ on In this case, the current compensation value of the switch branch in the on state is less than 0 and is greater than the negative value of the steady-state current value of the switch branch in the on state, that is, -i s <γ on <0.

[0092] Therefore, a current source close to steady state I is selected. ons The current compensation value is used to compensate the historical current source, so that the current value of the current source of the switch branch when switching the switch state is closer to the current value of the steady-state current source, thereby shortening the convergence time of the transient response. In addition, when the conduction steady state is reached, I ons =α on *G*v(t-△t)+β on *i(t-△t)+γ on =-i s , in -i s <γ on <0, α in the on-state on *G*v(t)+β on *i(t) s , thereby reducing the energy stored in the switch element in the on state, thereby reducing the virtual loss of the switch element.

[0093] In the off state, the steady-state current value of the switch branch is 0, that is, i s =0, then the steady-state value of the historical current source in the off state is I offs =G*v s , v s ​is the steady-state voltage value in the off state. Therefore, the expression of the historical current source I(t) shown in formula (6) is I(t) = α when it reaches the off steady state. off *G*v(t-△t)+β off *i(t-△t)+γ off =G*v s When the switch state switches from on to off, since v(t-△t)=0, the initial value of the historical current source after switching is I(t)=β off *i(t-△t)+γ off In this case, the current compensation value of the switch branch in the off state is greater than 0 and is less than the product of the steady-state voltage value of the switch branch in the off state and the admittance value of the switch branch, that is, 0<γ off <G*v s .

[0094] Therefore, a current source close to steady state I is selected. ons The current compensation value is used to compensate the historical current source, so that the current value of the current source of the switch branch when switching the switch state is closer to the current value of the steady-state current source, thereby shortening the convergence time of the transient response. In addition, when the off steady state is reached, I offs =α off *G*v(t-△t)+β off *i(t-△t)+γ off =G*v s , when 0<γ off <G*v s In the case of g, the α in the steady state is turned off. off *G*v(t)+β off *i(t) <G*v s , thereby reducing the energy stored in the switch element in the off state, thereby reducing the virtual loss of the switch element.

[0095] For example, see Figure 6 The comparison diagram of the state transition trajectory of the historical current source is shown in FIG. Figure 6 As shown, without introducing the constant term of the current compensation value, for example, I(t) = α*G*v(t-△t)+β*i(t-△t), the initial value of the historical current source after the switch state is switched is [I0, I1, I2, ..., I n ]0, n is the number of switching elements, the steady-state value of the historical current source is [I0, I1, I2, ..., I n ] S , the state transition trajectory is shown by the solid arrow. When the constant term of the current compensation value is introduced, the initial value of the historical current source after the switch state is switched is [I0, I1, I2, ..., In ] 0' The state transition trajectory is shown by the dotted arrow, and the time it takes for the switch branch to reach the steady-state value from the initial value is shortened.

[0096] In summary, the updated parameters of the historical current source are obtained through the above process (1)-(3). off and β on Determines the steady-state characteristics of the model corresponding to the switching element, α on and β off Determines the eigenvalue of the state matrix of the transient response, γ on and γ off Determines the initial value of the transient response. Under the premise of satisfying the steady-state characteristics and convergence, by configuring α on , α off , β on , β off , γ on and γ off It can accelerate dynamic response speed and reduce virtual loss.

[0097] After obtaining the circuit model, the host computer can send it to a programmable logic array (FPGA), which can be used to simulate the circuit model. The circuit model includes a relationship between the historical current source current values I(t) corresponding to each switch branch. The programmable logic array can update the historical current source current values at different simulation times based on the relationship between the current values I(t).

[0098] In an embodiment of the present application, a programmable logic array is used to update the current values of the current sources in the multiple switch branches at the current simulation moment according to the branch voltage values and branch current values of the multiple switch branches at the previous simulation moment during the simulation of the circuit model; and to update the branch current values and branch voltage values of the multiple switch branches at the current simulation moment according to the admittance values of the multiple switch branches and the current value of the current source at the current simulation moment.

[0099] Exemplarily, the programmable logic array is configured to, after obtaining branch voltage values and branch current values of the multiple switch branches at a previous simulation time, substitute the branch voltage values and branch current values of the multiple switch branches at the previous simulation time into the above formula (5), i.e., substitute I(t) = α*G*v(t-Δt) + β*i(t-Δt) + γ, to obtain the current values of the current sources in the multiple switch branches at the current simulation time. The voltage coefficient α, current coefficient β, and current compensation value γ corresponding to different switch branches may be the same or different.

[0100] After updating the current values of the current sources in the multiple switch branches at the current simulation time, the programmable logic array is configured to determine the input current value of each node of the circuit model based on the current values of the current sources in the multiple switch branches at the current simulation time; determine the node voltage value of each node based on the admittance values of the multiple switch branches and the input current value of each node; and update the branch current values and branch voltage values of the multiple switch branches at the current simulation time based on the node voltage values of each node. This makes the determined branch current values and branch voltage values more accurate.

[0101] Optionally, a programmable logic array is used to obtain an input column vector based on the current source of at least one switch branch at a current moment, the input column vector including n elements, where n is the number of nodes in the circuit model, a node refers to a connection point between at least two switch branches, and the value of element i indicates the input current value corresponding to node i, where i is a positive integer less than or equal to n; based on the node admittance matrix and the input column vector, the node voltage values corresponding to the n nodes are updated respectively, the node admittance matrix including n*n elements, the value of element ij indicates the admittance value between node i and node j, where j is a positive integer less than or equal to n.

[0102] For example, since the differential equation of each simulation element in the circuit model is equivalent to a circuit form in which the admittance is connected in parallel with a historical current source, Kirchhoff's law, an electrical theorem, is used to organize all simulation elements in the circuit topology into a linear equation system Hx = b. A general linear equation solver is then used to solve the linear equation system to achieve simulation solution. Here, H is the node admittance matrix, b is the node input current vector (i.e., the input column vector), and x is the node voltage vector. The node voltage vector x includes n elements, and the value of element j indicates the node voltage value corresponding to node j, where j is a positive integer less than or equal to n.

[0103] In an embodiment of the present application, the node admittance matrix can be obtained before the real-time simulation is run, and does not take up the time of each simulation step in the real-time simulation process. The solution of the linear equations is performed with each simulation step in the real-time simulation process. Optionally, the solution of the linear equations Hx=b is completed by the linear equation solver. For example, first obtain the inverse matrix H-1 of the node admittance matrix, and solve the node voltage value according to the following formula (17). Among them, the value of the node admittance matrix at each simulation time remains unchanged, the node voltage vector x[t] at the simulation time t is related to the input column vector b[t] at the simulation time t, and the input column vector b[t] at the simulation time t is affected by the historical current source, and then affected by the branch current value and branch voltage value at the previous simulation time t-△t.

[0104] x[t]=H -1 *b[t] Formula (17)

[0105] Therefore, each independent simulation component's impact on real-time simulation is concentrated in two aspects: the filling of the node admittance matrix before the simulation runs, and the updating of the input column vector b during the simulation run. The real-time simulation algorithm can be divided into two parts: simulation initialization before the real-time simulation starts, and real-time calculation during the simulation startup process. The filling of the node admittance matrix is completed during the simulation initialization process, while the updating of the element values of the input column vector is completed during the real-time calculation process.

[0106] Without additional reordering intervention, the node admittance matrix's elements are expanded in the order of node numbers. Optionally, the entire node admittance matrix formation process can be decoupled into a cumulative filling process of each simulated element's admittance value at the corresponding node position. Each simulated element independently contributes to the admittance matrix value at the corresponding position, with the filling being additive. For example, if a resistor element is associated with two nodes n0 and n1 in a circuit topology, it will have an impact at the intersection of rows n0 and n1, and columns n0 and n1, of the node admittance matrix. For elements in diagonal positions, the admittance value is positively filled, while for elements in non-diagonal positions, the admittance value is negatively filled. For elements connected to more than two nodes, such as three-phase elements, the same principles apply. If the three nodes connected to the three-phase element are consecutively numbered, the element is represented as a block matrix. For multiple elements connected to the same node, the admittance value of each element is independently added to the node admittance matrix. Each element only updates the admittance value associated with its node position and does not affect the admittance values at other positions.

[0107] For example, Figure 4 As an example of the circuit topology shown in the figure, after converting each component into a historical current source parallel admittance structure, the circuit topology is as follows Figure 7 As shown. Among them, the voltage source Us is converted into the structure of current source Is in parallel with resistor Rs, and the capacitor C is converted into the historical current source I hc Parallel admittance G c The structure of the inductor L1 is transformed into a historical current source I hl1 Parallel admittance G l1 The structure of the inductor L2 is transformed into a historical current source I h12 Parallel admittance G l2 structure. Figure 7 The node admittance matrix H of the circuit model shown is shown in the following formula (18). Taking the inductor L1 as an example, the admittance value G after the integral discretization of the inductor L1 is l1 It will affect the four elements of the node admittance matrix at the corresponding positions of node 1 and node 2.

[0108]

[0109] by Figure 7Taking node 2 as an example, based on the KCL law, the sum of the currents flowing into node 2 is equal to the sum of the currents output from node 2, so the KCL equation at node 2 can be: l1 *u1+(G l1 +G c +G l2 )*u2-G l2 *u3=-I hl1 +I c +I hl2 . Where v1 is the node voltage value at node 1, v2 is the node voltage value at node 2, and v3 is the node voltage value at node 3. Similarly, the corresponding KCL equations can also be listed for nodes 1 and 2. According to the KCL equations corresponding to different nodes, the linear equation group shown in the following formula (19) can be obtained. Where, Corresponding to the node voltage vector x[t] in formula (17), Corresponding to the node voltage vector b[t] in formula (17).

[0110]

[0111] Therefore, the historical current source of each branch affects the input column vector b. The historical current source of a simulation component is an iterative process of numerical integration and needs to be updated at every simulation moment. Each simulation component has its own historical current source. Updating the column vector elements associated with its node does not affect the column vector elements at other locations. For example, if the historical current source is associated with nodes 1 and 2 in the circuit topology, it will affect the first and second rows of the input column vector accordingly. According to the flow direction of the historical current source, the current flowing into the node is positively injected, and the current flowing out of the node is negatively injected.

[0112] From the filling process of the node admittance matrix, it can be seen that the node admittance matrix has positive definiteness, symmetry and sparsity. Positive definiteness: It is reflected in the fact that the parameters of the circuit elements are physical parameters and are almost positive numbers. Therefore, the diagonal elements of the node admittance matrix are positive numbers, and the off-diagonal elements are negative numbers. Therefore, the node admittance matrix is a positive definite matrix. Symmetry: It is reflected in the fact that the node admittance matrix is established based on Kirchhoff's law. Its elements are filled according to the node position. They are symmetrically distributed along the diagonal on the admittance matrix and have symmetry. Sparsity: It is reflected in the fact that the connection points of circuit elements in the topology are limited. It is rare for a node to have a traversal connection relationship with all nodes. Therefore, the probability of a 0 element appearing at the corresponding position of the node admittance matrix is very high. The node admittance matrix is sparse, and the larger the circuit scale, the more obvious the sparsity.

[0113] In summary, at each simulation moment, the current value of the historical current source of each branch is iteratively updated by the above formula (6), and the input column vector b[t] can be updated according to the updated current value of the historical current source of each branch, and then the node voltage value of each node is iteratively updated by the above formula (17). After updating the node voltage value of each node, the voltage across the element connecting the two nodes is the branch voltage. For example, if the branch connects node i and node j, and the node voltage values of node i and node j are Ui and Uj respectively, then the branch voltage value Uij of the branch is Ui-Uj. Then, based on the branch voltage value and admittance value of the branch, the branch current value of the branch can be calculated.

[0114] For example, the schematic diagram of the simulation process of the programmable logic array to the circuit model can be as follows: Figure 8 As shown. Among them, the historical current source update module is used to calculate the current value of the historical current source at the current simulation moment based on the branch current value, branch voltage value and the update relationship (i.e., component model) shown in formula (6) at the previous simulation moment. The input update module is used to determine the input voltage value at the current simulation moment through the voltage input item, and determine the input current value at the current simulation moment through the current input item. According to the current value of the historical current source at the current simulation moment and the input voltage value or input current value, the input column vector is updated. The node variable solving module is used to solve the node voltage value of each node according to the node variable relationship shown in formula (17). The branch current and voltage update module is used to calculate the branch voltage value and branch current value at the current simulation moment based on the node voltage value of each node. Thus, the above simulation cycle process is iteratively executed until the simulation task stops, for example, the simulation stop time is reached.

[0115] Optionally, the switch state processing module is configured to determine and switch the switch state of the switch branch. For example, the switch state processing module is connected to the control device and receives a drive signal from the control device to determine the switch state of the switch branch based on the drive signal. Optionally, the drive signal may be a pulse width modulation (PWM) signal. In this case, the historical current source update module is further configured to, in response to a switch state change in any switch branch, switch the update relational expression for the historical current source of any switch branch so that the updated relational expression after the switch change matches the switch state after the switch change.

[0116] In summary, in the simulation system provided by the embodiment of the present application, the switching elements in the circuit model are equivalent to parallel current sources and admittances, and the current value of the current source is updated at each simulation moment according to the branch voltage value and branch current value at the previous simulation moment, thereby avoiding virtual power loss during the switch state switching process. Moreover, in the process of updating the current value of the current source, not only the rate at which the switch branch converges to a steady state is increased by the voltage coefficient and the current coefficient, but also the initial state after the switch state is switched is made closer to the steady state value by the current compensation value. This effectively shortens the time for the switch branch to converge to a steady state after the switch state is switched, avoids transient errors caused by slow transient response, and improves the simulation accuracy of the simulation system.

[0117] The embodiment of the present application also provides a simulation method for a power conversion device. Optionally, the simulation method can be applied to the simulation system of the power conversion device provided in the above embodiment. Figure 9 As shown, the method includes but is not limited to step 901 and step 902.

[0118] Step 901 : Simulate a circuit model of a power conversion device, where the circuit model includes multiple switch branches, each switch branch is used to be equivalent to a switch element, and each switch branch includes a current source and an admittance connected in parallel.

[0119] Step 902, during the simulation of the circuit model, based on the branch voltage values and branch current values of the multiple switch branches at the previous simulation moment, the current values of the current sources in the multiple switch branches at the current simulation moment are updated; based on the admittance values of the multiple switch branches and the current values of the current sources at the current simulation moment, the branch current values and branch voltage values of the multiple switch branches at the current simulation moment are updated; wherein, the current value I(t) of the current source in each switch branch at the current simulation moment satisfies: I(t) = α*G*v(t-△t)+β*i(t-△t)+γ; α is the voltage coefficient, G is the admittance value of the switch branch, v(t-△t) is the branch voltage value at the previous simulation moment, β is the current coefficient, i(t-△t) is the branch current value at the previous simulation moment, γ is the current compensation value, and △t is the simulation step size.

[0120] In a possible implementation, the voltage coefficient, current coefficient, and current compensation value corresponding to the switch branch in the on state are different from the voltage coefficient, current coefficient, and current compensation value corresponding to the switch branch in the off state.

[0121] Optionally, the current coefficient of the switch branch in the on-state and the voltage coefficient in the off-state are values obtained by solving a discrete system model based on the final value theorem and steady-state characteristics. The discrete system model is a complex frequency domain relationship between the branch voltage and the branch current generated based on a relationship for the current value I(t). The steady-state characteristics include a steady-state voltage of 0 in the on-state and a steady-state current of 0 in the off-state. The voltage coefficient and the current coefficient of the switch branch in the on-state are coefficients within a range where the spectral radius is less than a threshold value. The spectral radius is the spectral radius of a state matrix of the current source iterated based on the relationship for the current value I(t).

[0122] In one possible implementation, the current compensation value of the switch branch in the on state is less than 0 and greater than the negative value of the steady-state current value of the switch branch in the on state; the current compensation value of the switch branch in the off state is greater than 0 and less than the product of the steady-state voltage value of the switch branch in the off state and the admittance value of the switch branch.

[0123] In an embodiment of the present application, the branch current values and branch voltage values of the multiple switch branches at the current simulation moment are updated according to the admittance values of the multiple switch branches and the current value of the current source at the current simulation moment, including: determining the input current value of each node of the circuit model according to the current value of the current source of the multiple switch branches at the current simulation moment, where a node refers to the connection point between two adjacent switch branches in the circuit model; determining the node voltage value of each node based on the admittance values of the multiple switch branches and the input current value of each node; and updating the branch current values and branch voltage values of the multiple switch branches at the current simulation moment according to the node voltage value of each node.

[0124] In a possible implementation, the admittance value G of the admittance of the switching branch satisfies:

[0125]

[0126] Where R is the damping resistance value of the switch branch in the off state, C is the capacitance value of the switch branch in the off state, L is the inductance value of the switch branch in the on state, and △t is the simulation step size.

[0127] For example, the schematic diagram of the simulation process of the power conversion device can be as follows: Figure 10As shown. First, 1. The circuit nodes are numbered and a netlist is formed. For example, the circuit nodes are numbered according to the circuit topology and component parameter information to form a circuit netlist. 2. Numerical integration discretization is equivalent to the electromagnetic transient model. For example, the backward Euler method is used to numerically integrate and discretize each component in the circuit and convert it into a unified current source parallel admittance structure. 3. Netlist analysis and initialization. For example, the circuit netlist is analyzed and initialized to obtain the node admittance matrix and its inverse matrix required for simulation, and the constant parameters required for simulation are obtained, such as the total simulation time, time step Δt, voltage coefficient α, current coefficient β, and current compensation value γ. 4. Node variable calculation. For example, after the simulation starts, the electromagnetic transient simulation loop iteratively calculates the input column vector and uses the product of the input column vector and the inverse matrix of the node admittance matrix to solve the node voltage value of each node. 5. Branch variable calculation. For example, based on the node voltage value obtained by calculating the node admittance matrix, the branch voltage value and branch current value of each branch are calculated. At this stage, all observable variables of the entire simulation system can be calculated. 6. Historical current source calculation. For example, the historical current source current values of each assistant are updated based on the branch voltage and branch current values of each branch for use in numerical integration discrete simulation iterations. 7. Simulation execution control: Determines whether the simulation has ended. If not, returns to the loop and executes steps 4-7 until the simulation task is terminated. After the simulation is completed, data such as the node voltages and branch currents recorded during the simulation process are output. The output data is analyzed, for example, by plotting voltage or current curves over time and calculating parameters such as power or energy to evaluate the performance and characteristics of the circuit.

[0128] In the embodiments of this application, Figure 9 For other implementations and beneficial effects of the simulation method for the power conversion device shown, please refer to the relevant content in the simulation system for the power conversion device mentioned above, and will not be repeated in the embodiments of this application.

[0129] In the embodiments of the present application, the terms "first", "second" and "third" are used for descriptive purposes only and should not be understood as indicating or implying relative importance. The term "at least one" means one or more, and "a plurality" means two or more.

[0130] In this application, the term "and / or" simply describes an association between related objects, indicating that three possible relationships exist. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this document generally indicates that the related objects are in an "or" relationship.

[0131] The above are merely optional embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and such modifications or substitutions should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A simulation system for a power conversion device, characterized in that: The system includes: a programmable logic array and a host computer; The host computer is used to send a circuit model of the power conversion device to the programmable logic array, wherein the circuit model includes a plurality of switch branches, each switch branch is used to be equivalent to a switch element, and each switch branch includes a current source and an admittance connected in parallel; The programmable logic array is configured to, during simulation of the circuit model, update the current values of the current sources in the multiple switch branches at a current simulation time based on the branch voltage values and branch current values of the multiple switch branches at a previous simulation time; and update the branch current values and branch voltage values of the multiple switch branches at a current simulation time based on the admittance values of the multiple switch branches and the current values of the current sources at the current simulation time; wherein the current value I(t) of the current source in each switch branch at the current simulation time satisfies: I(t)=α*G*v(t-△t)+β*i(t-△t)+γ; The α is the voltage coefficient, the G is the admittance value of the switching branch, the v(t-△t) is the branch voltage value at the last simulation moment, the β is the current coefficient, the i(t-△t) is the branch current value at the last simulation moment, the γ is the current compensation value, and the △t is the simulation step size.

2. The system according to claim 1, wherein: The voltage coefficient, current coefficient, and current compensation value corresponding to the switch branch in the on state are different from the voltage coefficient, current coefficient, and current compensation value corresponding to the switch branch in the off state.

3. The system according to claim 1 or 2, characterized in that The current coefficient of the switch branch in the on state and the voltage coefficient in the off state are values solved for the discrete system model based on the final value theorem and the steady-state characteristics. The discrete system model is a relationship between the branch voltage and the branch current in the complex frequency domain generated based on the relationship of the current value I(t). The steady-state characteristics include a steady-state voltage of 0 in the on state and a steady-state current of 0 in the off state.

4. The system according to any one of claims 1 to 3, characterized in that: The voltage coefficient of the switch branch in the on state and the current coefficient in the off state are coefficients within a range where the spectrum radius is less than a threshold value, and the spectrum radius is the spectrum radius of the state matrix of the current source iterated based on the relationship of the current value I(t).

5. The system according to any one of claims 1 to 4, characterized in that: The current compensation value of the switch branch in the on state is less than 0 and greater than the negative value of the steady-state current value of the switch branch in the on state; The current compensation value of the switch branch in the off state is greater than 0 and is less than the product of the steady-state voltage value of the switch branch in the off state and the admittance value of the switch branch.

6. The system according to any one of claims 1 to 5, characterized in that: The programmable logic array is configured to: determining an input current value of each node of the circuit model according to current values of the current sources of the multiple switch branches at a current simulation moment, wherein the node refers to a connection point between two adjacent switch branches in the circuit model; determining a node voltage value of each node based on the admittance values of the plurality of switch branches and the input current value of each node; The branch current values and branch voltage values of the plurality of switch branches at the current simulation moment are updated according to the node voltage values of the respective nodes.

7. The system according to any one of claims 1 to 6, characterized in that: The admittance value G of the admittance of the switch branch satisfies: Wherein, R is the damping resistance value of the switch branch in the off state, C is the capacitance value of the switch branch in the off state, and L is the inductance value of the switch branch in the on state.

8. A method for simulating a power conversion device, characterized in that: include: Simulating a circuit model of a power conversion device, wherein the circuit model includes a plurality of switch branches, each switch branch is used to be equivalent to a switch element, and each switch branch includes a current source and an admittance connected in parallel; During the simulation of the circuit model, updating the current values of the current sources in the multiple switch branches at the current simulation moment according to the branch voltage values and branch current values of the multiple switch branches at the previous simulation moment; The branch current values and branch voltage values of the multiple switch branches at the current simulation time are updated according to the admittance values of the multiple switch branches and the current values of the current sources at the current simulation time; wherein the current value I(t) of the current source in each switch branch at the current simulation time satisfies: I(t)=α*G*v(t-△t)+β*i(t-△t)+γ; The α is the voltage coefficient, the G is the admittance value of the switching branch, the v(t-△t) is the branch voltage value at the last simulation moment, the β is the current coefficient, the i(t-△t) is the branch current value at the last simulation moment, the γ is the current compensation value, and the △t is the simulation step size.

9. The method according to claim 8, characterized in that The voltage coefficient of the switch branch in the on state and the current coefficient in the off state are coefficients within a range where the spectrum radius is less than a threshold value, and the spectrum radius is the spectrum radius of the state matrix of the current source based on the iterative relationship of the current value I(t).

10. The method according to claim 8 or 9, characterized in that The current compensation value of the switch branch in the on state is less than 0 and greater than the negative value of the steady-state current value of the switch branch in the on state; the current compensation value of the switch branch in the off state is greater than 0 and less than the product of the steady-state voltage value of the switch branch in the off state and the admittance value of the switch branch.

Images