Process estimation system, process data estimation method, and recording medium
Provided are a process estimation system and a process data estimation method for appropriately estimating process data, and a program. The process estimation system includes: an input part configured to input actual sensor data detected by a sensor of a substrate processing apparatus; a virtual sensor data generation part configured to generate virtual sensor data for a virtual sensor based on the actual sensor data and a physical model; and a process data estimation part configured to estimate process data based on the virtual sensor data.
1 . A process estimation system comprising:
an input part configured to input actual sensor data detected by a sensor of a substrate processing apparatus;
a virtual sensor data generation part configured to generate virtual sensor data for a virtual sensor based on the actual sensor data and a physical model; and
a process data estimation part configured to estimate process data based on the virtual sensor data and a reaction model,
wherein the reaction model comprises:
a gas phase reaction model configured to estimate a decomposition amount of a precursor gas inside a processing space based on virtual sensor data of a gas pressure and virtual sensor data of a gas temperature inside the processing space; and
a surface reaction model configured to estimate an adsorption amount of the precursor gas adsorbed on a surface of a substrate as the process data, based on the decomposition amount of the precursor gas estimated by the gas phase reaction model and virtual sensor data of a temperature of the substrate.
2 . The process estimation system of claim 1 , wherein the process data estimation part is configured to estimate the process data based on the actual sensor data, the virtual sensor data and the reaction model.
3 . The process estimation system of claim 2 , wherein the sensor includes at least one of a temperature sensor, a gas flow rate sensor, a pressure sensor, a voltage sensor, a current sensor, and an emission spectroscopic analyzer.
4 . The process estimation system of claim 3 , wherein the physical model is a thermo-fluid model.
5 . The process estimation system of claim 4 , wherein the substrate processing apparatus is a film forming apparatus that performs a film forming process on the substrate, and
the process data is a thickness of a film formed on the substrate.
6 . The process estimation system of claim 5 , wherein the process data estimation part includes a machine learning part configured to perform a machine learning and generate a learned model by using a data set of the virtual sensor data and the process data as teacher data.
7 . The process estimation system of claim 6 , wherein the process data estimation part is configured to estimate the process data based on the virtual sensor data and the learned model.
8 . The process estimation system of claim 1 , wherein the sensor includes at least one of a temperature sensor, a gas flow rate sensor, a pressure sensor, a voltage sensor, a current sensor, and an emission spectroscopic analyzer.
9 . The process estimation system of claim 1 , wherein the physical model is a thermo-fluid model.
10 . The process estimation system of claim 1 , wherein the substrate processing apparatus is a film forming apparatus that performs a film forming process on the substrate, and
the process data is a thickness of a film formed on the substrate.
11 . The process estimation system of claim 1 , wherein the process data estimation part includes a machine learning part configured to perform a machine learning and generate a learned model by using a data set of the virtual sensor data and the process data as teacher data.
12 . A process data estimation method of estimating process data of a substrate processing apparatus, the process data estimation method comprising:
inputting actual sensor data detected by a sensor of the substrate processing apparatus;
generating virtual sensor data for a virtual sensor based on the actual sensor data and a physical model; and
estimating process data based on the virtual sensor data and a reaction model,
wherein the reaction model comprises:
a gas phase reaction model configured to estimate a decomposition amount of a precursor gas inside a processing space based on virtual sensor data of a gas pressure and virtual sensor data of a gas temperature inside the processing space; and
a surface reaction model configured to estimate an adsorption amount of the precursor gas adsorbed on a surface of a substrate as the process data, based on the decomposition amount of the precursor gas estimated by the gas phase reaction model and virtual sensor data of a temperature of the substrate.
13 . A non-transitory computer-readable recording medium that causes a computer to execute the process data estimation method of claim 12 .