TWI658349B - Process monitoring method and process monitoring system - Google Patents

Abstract

A process monitoring method includes the following steps: receiving set parameters; receiving process parameters within a specified time of detection during etching time; processing the process parameters every sampling time during the specified time of detection, and generating a process within a process sampling range Parameter feature code; judge the process compensation time, search for similar setting parameters according to the process parameter feature code to determine the process compensation time of the etching time; update the process parameter feature code to the database parameters. In addition, a process monitoring system has also been proposed.

Description

Process monitoring method and process monitoring system

The invention relates to a process monitoring method and a process monitoring system, and in particular to a process monitoring method and a process monitoring system that dynamically improve substrate etching time.

An electronic device such as a semiconductor element or a liquid crystal display device is formed by integrating a plurality of precise patterns. This pattern is formed by a plurality of process steps, including an etching step. There are two main types of etching in the microelectronic process, one is a wet etching method using a liquid etchant, and the other is a dry etching method using a plasma.

As far as the wet etching method is concerned, the control of the wet etching ability greatly affects the product quality and component performance. The conventional wet etching method is to set a fixed etching time by controlling the liquid concentration to accelerate or reduce to achieve the desired etching film thickness. However, the etching rate will change because the liquid concentration and the reaction temperature change with time, making the etching The process seems difficult to control. Therefore, how to provide a "process monitoring method and system" to solve the above-mentioned problems is a subject that needs to be solved in the related technical field.

An object of the present invention is to dynamically improve the etching time of a substrate to optimize the etching.

Another object of the present invention is to filter out the noise detected by the light sensing element due to pollution, moisture, reflection and other factors, so as to improve the quality of the received data and reduce the probability of judgment errors in subsequent calculations.

Another object of the present invention is to reduce the time difference due to the communication time of the light sensing element through the synchronization process, avoid time errors caused by measuring data at the same time point, and reduce the error rate in subsequent calculations.

An embodiment of the present invention provides a process monitoring method including the following steps: receiving a setting parameter, wherein the setting parameter includes a recipe parameter and a database parameter, the recipe parameter includes an etching time; a detection specification in the etching time Receive a process parameter within time, wherein the process parameter is generated from the light sensing element; process the process parameter every other sampling time during the detection specified time, and generate a corresponding process parameter characteristic code within a process sampling range; Determine the process compensation time, and search for similar setting parameters according to the feature code of the process parameter to determine a process compensation time for the etching time; update the feature code of the process parameter to the database parameter.

An embodiment of the present invention provides a process monitoring system, which includes a data processing unit, a process unit, and a PLC control unit. The data processing unit is used for receiving a setting parameter, the setting parameter includes a recipe parameter and a database parameter, and the recipe parameter includes an etching time. The process unit is used to transmit a process parameter. The PLC control unit signal connects the process unit and the data processing unit. The PLC control unit receives the process parameters, and the PLC control unit transmits the process parameters to the data processing unit. Among them, the data processing unit processes the process parameters within a specified time of detection in the etching time In order to generate the characteristic code of the corresponding process parameter, the data processing unit is based on the characteristic of the process parameter Code, search for similar setting parameters to determine the process compensation time instruction of the etching time, the data processing unit transmits the process compensation time instruction to the PLC control unit, and the PLC control unit controls the process compensation time of the process unit.

Based on the above, in the process monitoring method and the process monitoring system of the present invention, the detected process parameters are analyzed and processed within a specified time of the detection time in the etching time to determine whether to perform a process compensation time on the etching time and a substrate etching time. Make dynamic improvements.

100‧‧‧process monitoring system

110‧‧‧Processing unit

112‧‧‧light sensor

120‧‧‧PLC Control Unit

130‧‧‧Data Processing Unit

132‧‧‧Data Collection Module

134‧‧‧Noise Filter Module

136‧‧‧ Judgment Calculus Module

138‧‧‧Signal Control Module

140‧‧‧ human-machine interface unit

150‧‧‧Database Unit

S100‧‧‧Process Monitoring Method

S110 ~ S150‧‧‧step

S112 ~ S116‧‧‧step

S1142 ~ S1148‧‧‧step

S122 ~ S126‧‧‧step

S1262 ~ S1266‧‧‧step

S131 ~ S137‧‧‧ steps

S141 ~ S145‧‧‧step

FIG. 1 is a schematic diagram of a process monitoring system of the present invention.

FIG. 2 is a flowchart of a process monitoring method according to the present invention.

FIG. 3 is a flowchart of receiving setting parameters in FIG. 2.

FIG. 4 is a flowchart of processing recipe parameters in FIG. 3.

FIG. 5 is a flowchart of receiving process parameters in a specified time during the etching time detection in FIG. 2.

FIG. 6 is a flowchart of receiving a transmittance value in FIG. 5.

FIG. 7 is a flowchart of processing process parameters in FIG. 2.

FIG. 8 is a flowchart of determining a process compensation time in FIG. 2.

The specific embodiments of the present invention will be further described below with reference to the accompanying drawings and embodiments. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, but cannot limit the protection scope of the present invention.

FIG. 1 is a schematic diagram of a process monitoring system of the present invention. Please refer to Figure 1. The process monitoring system 100 of this embodiment is, for example, a system for monitoring an etching process. The process monitoring system 100 includes a process unit 110, a PLC control unit 120, a data processing unit 130, a human-machine interface unit 140, and a database unit 150.

The process unit 110 is used to process a substrate. The process unit 110 is, for example, an etching machine to etch the substrate. The process unit 110 is used to transmit a process parameter. The process unit 110 of this embodiment includes a plurality of light sensing elements 112. The process parameters are generated from the light sensing element 112, and the light sensing element 112 is used to detect the transmission of the substrate. Light transmittance to obtain the light transmittance value, and the corresponding process parameters can be analyzed and calculated through the light transmittance value. The light sensing element 112 is used to transmit the process parameters to the PLC control unit 120. The number of light sensing elements 112 in this embodiment For example, there are three, and the number and type of the light sensor components can be selected depending on the actual manufacturing process (such as the size of the substrate).

The PLC control unit 120 is a programmable logic controller. The PLC control unit 120 signals are connected to the process unit 110. The PLC control unit 120 receives the process parameters transmitted from the light sensing element 112. The PLC control unit 120 transmits the process parameters to Data processing unit 130.

The signal from the human-machine interface unit 140 is connected to the data processing unit 130. The human-machine interface unit 140 is used to display the user interface and allow the user to input the required recipe parameters to the data processing unit 130. The recipe parameters are, for example, etching solution, pressure, set temperature, substrate thickness, transfer speed, etching time, detection specified time, sampling time and other parameters. The etching time includes a continuous first time interval and a second time interval. Sampling time and detection specified time are defined in the second time interval, and the parameters of the recipe parameters can be adjusted depending on the actual machine and process requirements. The data of the database unit 150 is connected to the data processing unit 130, and the database unit 150 is used for storing and retrieving the database Parameters to the data processing unit 130.

The data processing unit 130 is configured to receive a setting parameter. The setting parameter includes a recipe parameter and a database parameter. The recipe parameter is received from the human-machine interface unit 140 and the database parameter is received from the database unit 150.

In detail, the data processing unit 130 includes a data collection module 132, a noise filter module 134, a judgment calculation module 136, and a signal control module 138. The signal of the data collection module 132 is connected to the noise filter module. 134. The signal of the noise filtering module 134 is connected to the judgment calculation module 136, and the signal of the judgment calculation module 136 is connected to the signal control module 138.

The signal of the data collection module 132 is connected to the PLC control unit 120. The data collection module 132 is used to receive the process parameters transmitted from the PLC control unit 120 within a specified period of time in the etching time. The noise filtering module 134 generates a final signal curve according to the process parameters and the database parameters, and the judgment calculation module 136 generates a corresponding code of the process parameters based on the final signal curve and the database parameters, and according to the feature codes of the process parameters After searching for similar setting parameters in the database unit 150, after searching for similar setting parameters, the database unit 150 can extract the corresponding feature code, and the corresponding feature code has a corresponding etching extension time, thereby determining the manufacturing process. Compensation time.

If the detection time specified in the etching time is reached, the judgment calculation module 136 transmits the calculated process compensation time instruction to the signal control module 138, and the signal control module 138 updates the feature code of the process parameter to the database unit 150. Database parameters, and the signal control module 138 is used to transmit the process compensation time instruction to the PLC control unit 120, and the PLC control unit 120 is used to control the process unit 110, so that the process unit 110 extends the required compensation time.

For example, the process unit 110 processes the substrate in the etching time, and the etching time For example, it is set to 120 seconds, wherein the first time interval is 40 seconds and the second time interval is 80 seconds. After the substrate is processed in the first time interval, the designated detection and specified time is 77 seconds in the second time interval. According to this, after 40 seconds have elapsed on the substrate of the process unit, before 77 seconds, the light sensing element 112 detects the transmittance of the substrate and transmits the process parameters to the PLC control unit 120. The PLC control unit 120 transmits the process parameters to the data processing unit 130.

If the detection time reaches 77 seconds, the judgment calculation module 136 in the data processing unit 130 processes the process parameters to determine the process compensation time. The signal control module 138 in the data processing unit 130 is used to transmit the process compensation time to the PLC control unit. 120. The process compensation time of the process unit 110 is controlled by the PLC control unit 120. The compensation time required by the process unit 110 is feedbacked, so that within 120 seconds of the etching time, it can be received and compensated within 77 seconds of the specified detection time. Subsequent required etching compensation time to dynamically improve substrate etching time. In addition, the signal control module 138 updates the feature code of the process parameter to the database parameter in the database unit 150.

FIG. 2 is a flowchart of a process monitoring method according to the present invention. FIG. 3 is a flowchart of receiving setting parameters in FIG. 2. FIG. 4 is a flowchart of processing recipe parameters in FIG. 3. FIG. 5 is a flowchart of receiving process parameters in a specified time during the etching time detection in FIG. 2. FIG. 6 is a flowchart of receiving a transmittance value in FIG. 5. FIG. 7 is a flowchart of processing process parameters in FIG. 2. FIG. 8 is a flowchart of determining a process compensation time in FIG. 2. The process monitoring method S100 of this embodiment can be used in the process monitoring system 100 shown in FIG. 1. The process monitoring method S100 includes the following steps S110 to S150.

Please refer to FIG. 2 first. First, step S110 is performed to receive a setting parameter. As shown in FIG. 1, the data processing unit 130 is configured to receive setting parameters. Step S110 may include the following steps S112 to S116. Referring to FIG. 3, step S112 is performed to receive recipe parameters. According to FIG. 1, the recipe parameters are received by the data collection module 132 in the data processing unit 130 from the human-machine interface unit. 140, for example, etching parameters, pressure, set temperature, substrate thickness, transfer speed, etching time, detection specified time, sampling time and other parameters, wherein the etching time includes a continuous first time interval and a second time interval, The sampling time and the detection specified time are the second time interval defined in the etching time, and the items for adjusting the recipe parameters can be adjusted according to the actual machine and process requirements.

Next, step S114 is performed to process the recipe parameters to generate a corresponding one of the reference characteristic curves. For details, please refer to FIG. 4. Step S114 includes the following steps S1142 to S1148, and step S1142 is performed to convert the recipe parameters into corresponding recipes. For the feature code, as shown in FIG. 1, the data collection module 132 transmits the recipe parameters to the noise filtering module 134, and the noise filtering module 134 converts the corresponding formula feature codes according to the recipe parameters. Then, step S1144 is performed, and the corresponding feature code in the database parameter is searched according to the formula feature code. In the case of FIG. 1, the noise filtering module 134 searches for the database parameter in the database unit 150 according to the formula feature code. Feature code.

Next, step S1146 is performed. If there is no corresponding feature code in the database parameters, the formula feature code is updated to the database parameters. In the case of FIG. 1, if the database unit 150 does not have a corresponding feature code, this formula feature The code may be the latest process, so the noise filtering module 134 can update the formula feature code to the database unit 150. Otherwise, step S1148 is performed. If there is a corresponding feature code in the database parameter, a corresponding reference feature curve is obtained according to the corresponding feature code in the database parameter. Therefore, if the noise filtering module 134 searches for the feature code corresponding to the formula feature code in the database unit 150, it can obtain the corresponding reference feature curve.

After obtaining the reference characteristic curve through the above steps S1142 to S1148, please refer to FIG. 3, and then proceed to step S116 to define a custom range sampling range, wherein the custom range sampling range is defined at the upper limit of the reference characteristic curve every sampling time. For example, the sampling time is 0.5 seconds, so every 0.5 seconds, the required process sampling range is defined on the reference characteristic curve.

Please refer to FIG. 2, and then proceed to step S120 to receive a process parameter within a specified time of detection in the etching time, wherein the process parameter is generated from the light sensing element. For details, please refer to FIG. 5. Step S120 includes the following steps S122 to S126, and step S122 is performed to process the substrate in the second time interval according to the recipe parameters. The set parameter includes a process parameter. The process parameter is generated in In the first time interval, as shown in FIG. 1, after the process unit 110 processes the substrate in the first time interval, the process parameters are generated. The PLC control unit 120 can receive the process parameters transmitted by the self-made process unit 110. The data collection module 132 can receive the process parameters transmitted from the PLC control unit 120, the data collection module 132 transmits the process parameters to the noise filter module 134, and the noise filter module 134 converts into the corresponding process feature code according to the process parameters Based on the process feature code, search for similar feature codes in the database unit 150.

Then step S124 is performed to detect the transmittance of the substrate to obtain a transmittance value. As shown in FIG. 1, the processing unit 110 continues to process the substrate in the second time interval. At the same time, the light sensing element 112 in the processing unit 110 detects the light transmittance of the substrate within a specified time to obtain light transmission. The numerical value of the transmittance can be analyzed to calculate the degree of film thickness of the substrate being etched by the transmittance value.

Next, step S126 is performed to receive a transmittance value to obtain a process parameter. For details, referring to FIG. 6, step S126 includes the following steps S1262 to S1266, and step S1262 is performed to set a plurality of receiving intervals. Next, step S1264 is performed to receive the transmittance values transmitted by the plurality of light sensing elements. In terms of FIG. 1, for example, the number of the light sensing elements 112 is three, and the three light sensing elements 112 respectively transmit the light transmittance. The light transmittance value is sent to the PLC control unit 120, and the data collection module 132 receives the light transmittance values corresponding to the three light sensing elements 112 and stores them.

Next, step S1266 is performed to combine the light transmittance values corresponding to each receiving interval into corresponding process parameters. For example, since the transmission time of each light sensing element 112 is different It is the same, so that the data collection module 132 receives the corresponding light transmittance value time. Therefore, the light transmittance values in a receiving interval are combined into corresponding process parameters through steps S1262 to S1266. In this way, the process parameters of the three light sensing elements 112 can be sent to the noise filtering module 134 synchronously, thereby reducing the gap in communication time due to the light sensing elements 112, and avoiding the time for measuring data at the same time point. The error can further reduce the error rate in subsequent analysis and calculation.

After receiving the process parameters in the above step S120, please refer to FIG. 2 again, and then proceed to step S130, process the process parameters every sampling time during the detection specified time, and generate the corresponding process parameter characteristic code within the process sampling range. For details, please refer to FIG. 7. Step S130 includes the following steps S131 to S137, and step S131 is performed to receive the process parameters at every sampling time. For FIG. 1, the sampling time is 0.5 seconds, so every 0.5 The second noise filtering module 134 receives the process parameters transmitted from the data collection module 132.

Next, step S132 is performed to generate a process feature code, where the process feature code generates a self-made process parameter. In terms of FIG. 1, the noise filtering module 134 converts process parameters into corresponding process feature codes. The process feature codes have, for example, polynomial function values, slopes, total integrals, or differences and have corresponding weight values. The code may be determined according to actual application and calculation requirements, and this embodiment does not limit this.

Next, step S133 is performed to search for a similar feature code in the database parameters according to the process feature code. As shown in FIG. 1, the noise filtering module 134 searches the database parameters of the database unit 150 for similar feature codes according to the process feature codes converted by the process parameters. Next, if there is no corresponding feature code in the database parameter, step S134 is performed to update the process feature code to the database parameter. Conversely, if there is a corresponding feature code in the database parameters, step S135 is performed. Generate a plurality of classification signal curves.

In terms of FIG. 1, if the database parameter in the database unit 150 does not have a corresponding feature code, the process feature code is updated to the database parameter in the database unit 150. Conversely, if the database parameter in the database unit 150 has a corresponding According to the feature code, a classification signal curve can be generated, and the number of classification signal curves can be determined according to the actual data collection.

Then step S136 is performed to obtain a final signal curve. The final signal curve is an average of a plurality of classified signal curves within the sampling range of the process. In terms of FIG. 1, the noise filtering module 134 extracts a classification signal curve that conforms to the sampling range of the process according to the customized range defined in step S116. This can be similar to the filtering process to filter out the light sensing element. 112 Noise detected due to pollution, moisture, reflection and other factors to improve the quality of received data and reduce the probability of judgment errors in subsequent analysis and calculation. After capturing the classification signal curves within the sampling range of the process, average these classification signal curves to obtain the final signal curve.

Then step S137 is performed to generate a feature code of the process parameter, wherein the feature code of the process parameter is generated from the final signal curve. In terms of FIG. 1, the noise filter module 134 converts the final signal curve into the characteristic code of the corresponding process parameter. The noise filter module 134 transmits the characteristic code of the process parameter to the database unit 150 and the judgment algorithm respectively. Group 136.

After obtaining the feature code of the process parameter, please refer to FIG. 2, and then proceed to step S140 to determine the process compensation time. According to the feature code of the process parameter, search for similar setting parameters to determine a process compensation time of the etching time. For details, referring to FIG. 8, step S140 includes the following steps S141 to S145, and step S141 is performed to determine whether the specified time is reached. Then step S142 is performed. If the specified time is detected, the step of judging the process compensation time is performed, and the process compensation time is transmitted.

In terms of FIG. 1, if 77 seconds of the specified time is detected, the judgment calculation module 136 processes the process parameters to determine the process compensation time. The judgment calculation module 136 searches for similar data in the database unit 150 according to the feature code of the process parameters. Set the parameters to determine the process compensation time. For example, after searching for the feature code of the process parameter, after finding a similar setting parameter, the corresponding feature code can be extracted from the database unit 150. The corresponding feature code has a corresponding etching extension time, so The corresponding feature code calculates and feedbacks how long the compensation is, and the judgment calculation module 136 transmits the process compensation time to the signal control module 138, and the signal control module 138 transmits the process compensation time instruction to the PLC control unit 120, the PLC control unit 120 The process compensation time instruction is transmitted to the process unit 110. Therefore, in this embodiment, through the above step S140, the etching compensation time required for the process unit 110 can be fed back to the specified time within the etching time, so as to dynamically improve the substrate etching time, instead of performing it after the substrate is etched. analysis. In addition, comparing the feature codes in the database unit 150 by generating the feature codes of the process parameters can reduce the amount of data and speed up subsequent search time. Then, as shown in FIG. 2, step S150 is performed to update the feature code of the process parameter to the database parameter. Based on this, the signal control module 138 updates the feature code of the process parameter to the database parameter in the database unit 150.

In addition, if the judgment calculation module 136 judges that no process compensation time is required through the above step S140, the judgment calculation module 136 transmits a command to remove the substrate to the signal control module 138, and the signal control module 138 transmits a command to remove the substrate to the PLC control The unit and the PLC control unit receive the instruction for removing the substrate transmitted by the signal control module 138. The PLC control unit transmits the instruction for removing the substrate to the processing unit 110. After the processing unit 110 receives the instruction for removing the substrate, the substrate is removed.

Please refer to FIG. 8 again. If the detection time is not reached, step S143 is performed to determine whether it is within the sampling time range. If it is not within the sampling time, step S144 is performed, and step S144 is executed. In step S110, the step S112 of receiving the recipe parameters is performed. If the sampling time is within the sampling time range, the step S145 is performed, and the step S120 is performed to receive the step S124 of detecting the light transmittance of the substrate in the process parameters.

In summary, in the process monitoring method and the process monitoring system of the present invention, the detected process parameters are analyzed and processed within a specified time of the detection in the etching time to determine whether to perform a process compensation time on the etching time and to control the substrate. The etching time is dynamically improved.

Furthermore, the present invention captures a classification signal curve that fits within the sampling range of the process. This can be similar to a filtering process to filter out noise detected by light sensing elements due to pollution, moisture, reflection and other factors. , In order to improve the quality of the received data and reduce the probability of judgment errors in subsequent calculations.

In addition, the present invention synchronously sends the process parameters of different light sensing elements, thereby reducing the gap in communication time due to the light sensing elements, which can also avoid the time error of measuring the data at the same time point, and can reduce subsequent calculations. Error rate.

The above description only describes the preferred embodiments or examples of the technical means adopted by the present invention to solve the problem, and is not intended to limit the scope of patent implementation of the present invention. That is, all changes and modifications that are consistent with the meaning of the scope of patent application of the present invention, or made according to the scope of patent of the present invention, are covered by the scope of patent of the present invention.

Claims (15)

A manufacturing process monitoring method includes the following steps: receiving a setting parameter, wherein the setting parameter includes a recipe parameter and a database parameter, the recipe parameter includes an etching time; and receiving a detection time within a specified time of the etching time. Process parameters, wherein the process parameters are generated from light sensing elements; the process parameters are processed at every sampling time during the specified time of the detection, within a process sampling range to generate corresponding feature code of the process parameters; Determine the process compensation time, search for similar setting parameters according to the feature code of the process parameter to determine a process compensation time of the etching time; and update the feature code of the process parameter to the database parameter. The process monitoring method according to item 1 of the scope of patent application, wherein the step of receiving the set parameter includes the following steps: receiving the recipe parameter; processing the recipe parameter to generate a corresponding one of the reference characteristic curves; and defining the process Sampling range, wherein the sampling range of the process is defined on the reference characteristic curve every sampling time. The process monitoring method described in item 2 of the scope of patent application, wherein the step of processing the recipe parameters includes the following steps: converting the recipe parameters into corresponding recipe feature codes; searching the database parameters according to the recipe feature codes Corresponding feature code in the database parameter; if there is no corresponding feature code in the database parameter, update the formula characteristic code to the database parameter; and if there is a corresponding feature code in the database parameter, according to the corresponding parameter in the database parameter To obtain the corresponding reference characteristic curve. The process monitoring method according to item 1 of the scope of patent application, wherein the etching time includes a continuous first time interval and a second time interval, and the process parameter is received within the specified time of the detection in the etching time. The step includes the following steps: manufacturing a substrate in the second time interval according to the recipe parameters, wherein the setting parameter includes a processed parameter generated in the first time interval; and detecting the substrate Transmittance to obtain the transmittance value; receive the transmittance value to obtain the process parameters. The process monitoring method according to item 4 of the scope of patent application, wherein the step of receiving the light transmittance value includes the following steps: setting a plurality of receiving intervals; receiving the light transmittance value transmitted by the plurality of light sensing elements; And combining the light transmittance values corresponding to each of the receiving intervals into corresponding process parameters. The process monitoring method according to item 1 of the scope of patent application, wherein the step of processing the process parameter every sampling time in the detection specified time includes the following steps: receiving the process parameter every sampling time ; Generate a process feature code, wherein the process feature code is generated from the process parameters; according to the process feature code, search for a similar feature code in the database parameter; if there is no corresponding feature code in the database parameter, The process feature code is updated to the database parameter; if there is a corresponding feature code in the database parameter, a plurality of classification signal curves are generated, wherein the plurality of classification signal curves are generated from the feature code; a final signal curve is obtained, The final signal curve is an average from the plurality of classification signal curves within the sampling range of the process; a feature code of the process parameter is generated, wherein the feature code of the process parameter is generated from the final signal curve. The process monitoring method according to item 1 of the scope of patent application, wherein the step of judging the process compensation time includes the following steps: judging whether the detection specified time is reached; if the detection specified time is reached, performing the step of judging the process compensation time To transmit the process compensation time; if the detection specified time is not reached, determine whether it is within the sampling time range; if it is not within the sampling time range, perform the step of receiving the set parameter; and if the sampling time is reached, execute Steps to receive the process parameters. A process monitoring system includes: a data processing unit for receiving a setting parameter, the setting parameter including a recipe parameter and a database parameter, the recipe parameter including an etching time, and a process unit for transmitting a process parameter And a PLC control unit, a signal connecting the process unit and the data processing unit, the PLC control unit receives the process parameters, and the PLC control unit transmits the process parameters to the data processing unit, one of which is in the etching time Detecting the specified time, the data processing unit processes the process parameters to generate a corresponding feature code of the process parameter. The data processing unit searches for similar setting parameters according to the feature code of the process parameter to determine the etching time. A process compensation time instruction, the data processing unit transmits the process compensation time instruction to the PLC control unit, and the PLC control unit controls the process compensation time of the process unit. The process monitoring system according to item 8 of the scope of patent application, further comprising: a database unit connected to the data processing unit by a signal, the database unit being used to store and extract the database parameters to the data processing unit. The process monitoring system according to item 9 of the scope of patent application, wherein the data processing unit includes a noise filter module, and the signal of the noise filter module is connected to the database unit, and the noise filter module is based on the process parameters. And the database parameters to generate a final signal curve. The process monitoring system described in item 10 of the scope of patent application, wherein the data processing unit includes a judgment calculation module, and the judgment calculation module signal is connected to the noise filtering module signal and the database unit, and the judgment calculation The module generates a corresponding feature code of the process parameter according to the final signal curve and the database parameter. The process monitoring system according to item 11 of the scope of patent application, wherein the data processing unit includes a signal control module, and the signal control module signals connect the database unit and the PLC control unit, and the judgment calculation module is used for The process compensation time instruction is transmitted to the signal control module. If the detection specified time in the etching time is reached, the signal control module is used to transmit the process compensation time instruction to the PLC control unit, and the signal control module The group updates the feature code of the process parameter to the database parameter in the database unit. The process monitoring system according to item 10 of the scope of patent application, wherein the data processing unit includes a data collection module, the PLC control unit signals are connected to the data collection module, and the PLC control unit transmits the process parameters to the data collection Module, the data collection module signal is connected to the noise filter module, and the data collection module transmits the process parameters to the noise filter module. The process monitoring system according to item 8 of the scope of patent application, wherein the PLC control unit transmits the process parameters to the data processing unit, the process unit includes a plurality of light sensing elements, and the process parameters are generated from the plurality of light The sensing element is used for detecting the light transmittance of the substrate. The process monitoring system according to item 8 of the scope of patent application, further comprising: a human-machine interface unit, the signal is connected to the data processing unit, and the human-machine interface unit is used for a user to input the recipe parameters to the data processing unit.