Multi-physics co-simulation method of power semiconductor modules

A multi-physics co-simulation method of a power semiconductor module, comprising the following steps:

• step 1, predefining parameters, performing a circuit simulation under a given initial temperature in a PSpice, and suspending circuit simulation when a preset step length of data exchange is reached;
• step 2, calculating a power loss data by adopting a power loss calculation method in a MATLAB, and transmitting power loss data into a COMSOL;
• step 3, performing a heat-force simulation in the COMSOL, and suspending heat-force simulation when the preset step length of data exchange is reached;
• step 4, feeding back an extracted parameter of a junction temperature to a circuit model of the PSpice by the MATLAB;
• and taking the parameter of junction temperature as a new initial value, and continuing to perform simulation of a next time step length of data exchange based on a former state;
• step 5, taking a state when a former loop is terminated as the initial state of a next loop, iterating to perform dynamic continuous multi-physics simulation, and terminating co-simulation when an iteration termination condition is met;
• wherein the multi-physics co-simulation method of the power semiconductor module further comprises:
• enabling a time step length of data exchange of the PSpice and the COMSOL to be equal to the time step length of data exchange of the COMSOL and the PSpice;
• and setting different time step lengths of data exchange to control a rate of the circuit simulation and a rate of the heat-force simulation, and performing multi-rate multi-physics coupling simulation;
• the step of setting the different time step lengths of data exchange comprises the following steps:
• constructing a first-order variable step length three-point numerical differentiation formula based on a Lagrange function by extracting and iterating numerical solutions of temperature coupling state variables of nodes, and calculating and obtaining a first-order derivative of an interpolation function of a current node:
• L
• n
• ′
• (
• t
• i
• )
• =
• λ
• i
• λ
• i
• -
• 1
• (
• λ
• i
• -
• 1
• +
• λ
• i
• )
• ⁢
• T
• ⁡
• (
• t
• i
• -
• 2
• )
• -
• λ
• i
• -
• 1
• +
• λ
• i
• λ
• i
• -
• 1
• ⁢
• λ
• i
• ⁢
• T
• ⁡
• (
• t
• i
• -
• 1
• )
• +
• λ
• i
• -
• 1
• +
• 2
• ⁢
• λ
• i
• λ
• i
• (
• λ
• i
• -
• 1
• +
• λ
• i
• )
• ⁢
• t
• ⁡
• (
• t
• i
• )
• wherein ti−2, ti−1, ti represent three adjacent continuous data exchange time points;
• ti represents a current time node;
• the relational expressions:
• ti−1=ti−2+λi−1, and ti=ti−2+λi−1+λi are met;
• T(ti−2), T(ti−1) and T(ti) represent corresponding numerical solutions of temperature at ti−2, ti−1, ti;
• and λi and λi−1 represent time step lengths;
• a step length adjustment strategy is established according to the first-order derivative of the interpolation function of the current node;
• and a step length decision interval is set as [ε1, ε2], and step length adjustment coefficients are set as a and b respectively, wherein a∈(0,1), b∈(1,+∞), and the following conditions are met:
• (1) if |L′n(ti)|∈[ε1, ε2], λi+1=λi;
• (2) if |L′n(ti)|≤ε1, λi+1=λi×b;
• and a time step length of data exchange is appropriately increased;
• and
• (3) if |L′n(ti)|≥ε2, λi+1=λi×a;
• and a step length is decreased.