Virtual sensors for combustion furnaces

Abstract

A method for operating a furnace includes the steps of: providing a furnace comprising a feed material, an oxygen-containing flow, and a fuel-containing flow; providing a virtual sensor comprising a model generated using inputs comprising condition data collected by a physical gas sensor pre-placed in the furnace; setting up the virtual sensor to communicate with the furnace; and operating the furnace by a combustion reaction, wherein during the operation of the furnace, the virtual sensor receives further condition data, and in response to the receipt of further condition data, the virtual sensor transmits a signal to the furnace to use the model to determine the combustion products based on the further condition data, and to automatically adjust the flow rate of at least one input based on the determined combustion products.

Description

[Technical Field]

[0001] Cross-reference of related applications This application claims priority to U.S. Utility Patent Application No. 18 / 660,630, filed on 10 May 2024, and U.S. Provisional Patent Application No. 63 / 469,695, filed on 30 May 2023, the entire disclosure of which is incorporated herein by reference.

[0002] This disclosure relates to a method for operating a furnace, and in some non-limiting embodiments or aspects, to a method for operating a furnace using virtual sensors. This disclosure also relates to a furnace system and virtual sensors. [Background technology]

[0003] Combustion furnaces often use physical sensors placed inside the furnace to monitor operating conditions. However, given the harsh conditions present inside the furnace (e.g., extreme temperature conditions), certain sensors exposed to these conditions will rapidly degrade and / or fail. For example, sensors for measuring oxygen composition may rapidly degrade or fail inside a combustion furnace. These sensors can be expensive, and consequently, the short lifespan of these sensors caused by harsh combustion furnace conditions increases the cost of the manufacturing process. [Overview of the Initiative]

[0004] According to some non-limiting aspects of the present disclosure, a method for operating a furnace using a virtual gas sensor includes the steps of: providing a furnace including a first input section containing a supply material configured to be heated in the furnace; a second input section containing an oxygen-containing flow; and a third input section containing a fuel-containing flow; and providing a virtual sensor including a model for determining combustion products in the furnace, wherein the model is generated using inputs including condition data, the condition data including operating conditions of the furnace collected by a physical gas sensor pre-placed in the furnace; and configuring the virtual sensor to receive further condition data relating to the operating conditions of the furnace. The steps include: installing a sensor to communicate with a furnace; and operating the furnace by a combustion reaction generated by combining oxygen from an oxygen-containing stream and fuel from a fuel-containing stream, wherein the combustion reaction heats the supply material in the furnace; and during the operation of the furnace, the virtual sensor receives further condition data; and in response to the receipt of further condition data, the virtual sensor uses a model to determine combustion products based on the further condition data, and based on the determined combustion products, the virtual sensor transmits a signal to the furnace to automatically adjust the flow rate of at least one of a first input, a second input, and a third input.

[0005] In some non-limiting embodiments, the method may further include the steps of: placing a physical gas sensor in a furnace; operating the furnace by a combustion reaction during a training period before providing a virtual sensor; collecting conditional data using the physical gas sensor during the training period; and generating a model using the conditional data collected during the training period. The combustion products are oxygen, CO, CO2, NO xThe supply material may include at least one of the following: , and unburned fuel composition. The supply material may include glass batch material. Physical sensors may be placed in the furnace crown and / or accumulator to collect conditional data, and the readings of the physical gas sensors may not be inputs for determining the combustion products in the furnace at any given point in time during furnace operation. Conditional data may include at least one of the following: fuel flow rate, oxygen flow rate, air flow rate, furnace temperature, oxygen composition, carbon dioxide composition, and carbon monoxide composition. Further conditional data may include oxygen, CO, CO2, NO x The virtual sensor does not necessarily have to contain at least one composition of the unburned fuel, and based on further conditional data, it detects oxygen, CO, CO2, NO x The model can determine the composition of at least one of the unburned fuels. The model can be generated based on conditional data and reaction chemistry. The signal sent to the reactor can automatically adjust the flow rate of the oxygen-containing flow into the reactor.

[0006] In some non-limiting embodiments, the method may further include the steps of: activating physical sensors in a furnace while a virtual sensor receives further conditional data during furnace operation; determining measured combustion products in the furnace using the physical sensor; and comparing the measured combustion products with those determined by the virtual sensor to determine a deviation. The method may further include updating a model using the measured combustion products in response to the deviation exceeding a threshold. At some point during furnace operation, the furnace may include at least one of a fuel flow sensor, an oxygen flow sensor, an air flow sensor, and a temperature sensor, and the furnace does not use readings from a physical gas sensor as input for determining combustion products in the furnace. The method may further include the step of providing a plurality of virtual sensors, including a first virtual sensor including a first model and a second virtual sensor including a second model, wherein the first model is generated using an input including conditional data relating to operating conditions at a first location in the furnace, and the second model is generated using an input including conditional data relating to operating conditions at a second location in the furnace.

[0007] According to some non-limiting aspects of the present disclosure, a furnace system includes a furnace comprising a first input containing a feed material configured to be heated within the furnace, a second input containing an oxygen-containing flow, and a third input containing a fuel-containing flow, and configured to operate by a combustion reaction produced by combining oxygen from the oxygen-containing flow and fuel from the fuel-containing flow; and a virtual sensor comprising a model for determining combustion products within the furnace, the model being generated using an input containing condition data, the condition data comprising operating conditions of the furnace collected by a physical gas sensor pre-placed within the furnace, the virtual sensor communicating with the furnace so as to be configured to receive further condition data relating to the operating conditions of the furnace, the virtual sensor being configured to determine the combustion products of the furnace in response to the receipt of further condition data, and the virtual sensor being configured to transmit signals to the furnace to automatically adjust the flow rate of at least one of the first input, the second input, and the third input based on the determined combustion products.

[0008] In some non-limiting embodiments, the system may further include at least one of a fuel flow sensor, an oxygen flow sensor, an air flow sensor, and a temperature sensor located within the furnace. At some point during furnace operation, the furnace system may not use readings from physical gas sensors as input for determining the combustion products within the furnace. The combustion products within the furnace are oxygen, CO, CO2, NO x The supply material may include at least one of the following: , and unburned fuel composition. The supply material may include glass batch material. The condition data may include at least one of the following: fuel flow rate, oxygen flow rate, air flow rate, furnace temperature, oxygen composition, carbon dioxide composition and carbon monoxide composition. Further condition data may include oxygen, CO, CO2, NO x The virtual sensor does not necessarily have to contain at least one composition of the unburned fuel, and based on further conditional data, it detects oxygen, CO, CO2, NO xand determine the composition of at least one of the unburned fuels. The model can be generated based on conditional data and reaction chemistry. The signal sent to the reactor can be configured to automatically adjust the flow rate of oxygen-containing flow into the reactor.

[0009] In some non-limiting embodiments, the model can be configured to activate physical sensors in the furnace while the virtual sensors receive further conditional data, to use the physical sensors to determine the measured combustion products in the furnace, to compare the measured combustion products with the combustion products determined by the virtual sensors to determine the deviation, and to update the model with the measured combustion products in response to the deviation exceeding a threshold. A plurality of virtual sensors can be provided, including a first virtual sensor containing a first model and a second virtual sensor containing a second model, the first model being generated using an input containing conditional data relating to the operating conditions of a first location in the furnace, and the second model being generated using an input containing conditional data relating to the operating conditions of a second location in the furnace.

[0010] According to some non-limiting aspects of the present disclosure, the virtual sensor includes at least one processor that stores a model for determining combustion products in a furnace, the model being generated using inputs including conditional data, the conditional data including furnace operating conditions collected by physical gas sensors pre-placed in the furnace.

[0011] In some non-limiting embodiments, the furnace can include a first input section containing a feedstock configured to be heated within the furnace, a second input section containing an oxygen-containing stream, and a third input section containing a fuel-containing stream, and the furnace is configured to operate by a combustion reaction generated by combining oxygen from the oxygen-containing stream and fuel from the fuel-containing stream. At least one processor is programmed or configured to communicate with the furnace to receive further condition data related to the operating conditions of the furnace, to determine the combustion products of the furnace in response to the receipt of the further condition data, and to transmit a signal to the furnace to automatically adjust the flow rate of at least one of the first input section, the second input section, and the third input section based on the determined combustion products. At a certain point during the operation of the furnace, the furnace system may not use the readings of the physical gas sensors as inputs for determining the combustion products within the furnace. The combustion products within the furnace can include at least one of oxygen, CO, CO2, NO x , and unburned fuel composition. The feedstock can include a glass batch material. The condition data can include at least one of fuel flow rate, oxygen flow rate, air flow rate, furnace temperature, oxygen composition, carbon dioxide composition, and carbon monoxide composition. The further condition data may not include the composition of at least one of oxygen, CO, CO2, NO x , and unburned fuel, and at least one processor is configured to determine the composition of at least one of oxygen, CO, CO2, NO x , and unburned fuel based on the further condition data. The model can be generated based on the condition data and reaction chemistry. The signal transmitted to the furnace can be configured to automatically adjust the flow rate of the oxygen-containing stream to the furnace. At least one processor is further programmed or configured to receive the measured combustion products within the furnace determined by the physical sensors within the furnace while the virtual sensor is receiving the further condition data, to determine the deviation by comparing the measured combustion products with the combustion products determined by the virtual sensor, and to update the model based on the measured gas content in response to the deviation exceeding a threshold value.

[0012] According to some non-limiting aspects of the present disclosure, the plurality of virtual sensors described herein form a system, and the plurality of virtual sensors include a first virtual sensor including a first model and a second virtual sensor including a second model. The first model is generated using an input including condition data related to the operating conditions at a first position within the furnace, and the second model is generated using an input including condition data related to the operating conditions at a second position within the furnace.

[0013] The present disclosure is described with reference to the following drawings, and like reference numerals identify like parts throughout.

Brief Description of the Drawings

[0014] [Figure 1] It is a process flow diagram of a method for operating a furnace by an existing system. [Figure 2] It is a process flow diagram of a method for operating a furnace using virtual sensors according to some aspects of the present disclosure. [Figure 3] It is a process flow diagram of a method for generating virtual sensors according to some aspects of the present disclosure. [Figure 4] It is a schematic diagram of a furnace system by an existing system. [Figure 5] It is a schematic diagram of a furnace system having virtual sensors according to some aspects of the present disclosure. [Figure 6] It is a schematic diagram of a furnace system having virtual sensors according to some aspects of the present disclosure. [Figure 7] It is a schematic diagram of a model training system according to some aspects of the present disclosure. [Figure 8] It is a schematic diagram of a model implementation system according to some aspects of the present disclosure. [Figure 9] It is a schematic diagram of a model recalibration system according to some aspects of the present disclosure. [Figure 10] It is a schematic diagram of a modeling process according to some aspects of the present disclosure. [Figure 11] This figure shows test data from a virtual sensor compared to a physical sensor, according to several aspects of this disclosure. [Figure 12A] This figure shows test data from a virtual sensor compared to a physical sensor, according to several aspects of this disclosure. [Figure 12B] This figure shows test data from a virtual sensor compared to a physical sensor, according to several aspects of this disclosure. [Modes for carrying out the invention]

[0015] Where used herein, spatial or directional terms such as “left,” “right,” “inside,” “outside,” “up,” and “down” are relevant to this disclosure as shown in the drawings. However, it should be understood that this disclosure may assume various alternative orientations, and therefore such terms should not be considered limiting. Furthermore, where used herein, all figures representing dimensions, physical properties, processing parameters, amounts of components, reaction conditions, etc., used herein and in the claims should be understood to be modified in all cases by the term “approximately.” Therefore, unless otherwise indicated, the figures described in the following specification and claims may vary depending on the desired properties to be obtained by this disclosure. At the very least, and not as an attempt to limit the application of the doctrine of equivalents to the claims, each figure should be interpreted by taking into account the number of significant figures reported and applying ordinary rounding techniques. Furthermore, all scopes disclosed herein should be understood to include the start and end scope values ​​and any sub-scopes contained therein. For example, a range written as "1 to 10" should be considered to include all subranges between (and including) the minimum value of 1 and the maximum value of 10, i.e., all subranges that start with a minimum value of 1 or greater and end with a maximum value of 10 or less, such as 1 to 3.3, 4.7 to 7.5, 5.5 to 10, etc. "One (a)" or "one (an)" refers to one or more.

[0016] As used herein, the terms “communicate” and “communicate” may refer to the reception, acceptance, transmission, forwarding, provision, etc., of information (e.g., data, signals, messages, instructions, commands, etc.). When one unit (e.g., a device, a system, a component of a device or system, a combination thereof) communicates with another unit, it means that one unit can directly or indirectly receive information from and / or directly or indirectly transmit (e.g., transmit) information to the other unit. This may refer to direct or indirect connections that are essentially wired and / or wireless. Furthermore, two units can communicate with each other even if the transmitted information is modified, processed, relayed, and / or routed between the first and second units. For example, the first unit can communicate with the second unit even if the first unit passively receives information and does not actively transmit information to the second unit. As another example, the first unit can communicate with the second unit if at least one intermediate unit (for example, a third unit located between the first and second units) processes information received from the first unit and transmits the processed information to the second unit. In some non-limiting embodiments or aspects, a message may refer to a network packet containing data (for example, a data packet).

[0017] Furthermore, all documents referenced herein, including, but not limited to, issued patents and patent applications, should be considered as being "incorporated by reference" in their entirety.

[0018] This disclosure relates to a combustion furnace used in an industrial process. The industrial process may be any industrial process in which an air and / or oxygen combustion furnace can be used. For example, the furnace may be an air and / or oxygen combustion furnace used to manufacture glass.

[0019] The following is a non-limiting background description of a furnace suitable for use in this disclosure.

[0020] The furnace may be a large-capacity furnace. For example, the furnace may be a glassmaking furnace. The furnace may include a first end or opening into which raw materials can be supplied. Once inside the furnace, the raw materials are heated to melt and form a molten material. The molten material may flow through a second end or discharge end.

[0021] A furnace may have a combustion chamber and a molten tank. The molten tank may communicate with a feeder. The molten tank may be where the raw material is melted. The feeder may hold unmolten raw material and supply unmolten raw material to the molten tank. Above the molten tank in the furnace, there may be a combustion chamber. The combustion chamber may comprise one or more burners that supply and burn oxygen-containing streams and fuel-containing streams (e.g., carbon-based fuels), thereby supplying heat to melt the material in the molten tank. The combustion chamber and molten tank may form a molten apparatus crown as shown and described herein.

[0022] In one non-limiting example, the raw material may be raw glass material (also called "glass batch material"). The raw glass material can be placed in the furnace at a first end or opening to the furnace by a charging device or feeder. Inside the furnace, the raw glass material can be melted to form molten glass. The molten glass can flow out of the outlet into the refining zone.

[0023] Burners can be positioned in openings in the side walls of the furnace. A furnace typically has at least two side walls, a first side wall and a second side wall, with the first side wall opposite the second side wall. The side walls may have openings configured to receive burners. Burners can be configured to supply oxygen-containing gas, fuel, and / or a mixture of oxygen-containing gas and fuel to the furnace, where the fuel burns to form a flame. The flame from the burner provides energy for melting the raw materials. Burners can extend through the walls of the furnace or through the ceiling of the furnace. A furnace may have burners in one side wall, or in the first and second side walls. A furnace may have at least four, at least six, at least eight, at least ten, or at least twelve burners, and / or up to 30 burners, up to 24 burners, up to 20 burners, or up to 16 burners. The burners on the first sidewall can be staggered (as opposed to being aligned) with the burners on the second sidewall.

[0024] The furnace may be an air and / or oxygen combustion furnace. Burners may be installed on the side walls at a certain distance from the surface of the molten material to provide an appropriate distribution of energy for melting the raw glass material. For example, the burners may be at least 0.25 m, or at least 0.40 m, and less than 1.0 m or 0.8 m from the surface of the molten material.

[0025] Oxygen-containing gases can include pure oxygen, air, or other oxygen-nitrogen gas blends.

[0026] Fuel-containing streams can contain any hydrocarbons commonly burned in industrial furnaces. Non-limiting examples of fuel-containing streams include natural gas, fuel oil, coke, coal, or diesel.

[0027] During the combustion process, exhaust gases may form inside the combustion chamber. These exhaust gases contain NO x May contain gas. NO xThe gas may be formed by the reaction of oxygen with nitrogen gas. This reaction may occur due to the high temperature present inside the furnace.

[0028] While furnaces used in glass manufacturing are described herein, it will be understood that furnaces used for other purposes, such as those used in metallurgical applications, those used for carrying out chemical reactions, and those used for hydrocarbon refining, are also within the scope of this disclosure.

[0029] This disclosure relates to a method for operating a furnace, such as a furnace having any of the aforementioned components, using a virtual gas sensor. The method includes the steps of: providing a furnace comprising a first input section containing a supply material configured to be heated in the furnace; a second input section containing an oxygen-containing flow; and a third input section containing a fuel-containing flow; providing a virtual sensor including a model for determining combustion products in the furnace, wherein the model is generated using inputs containing condition data, the condition data including furnace operating conditions collected by a physical gas sensor pre-placed in the furnace; installing the virtual sensor to communicate with the furnace so that it is configured to receive further condition data relating to the furnace operating conditions; and operating the furnace by a combustion reaction produced by combining oxygen from the oxygen-containing flow and fuel from the fuel-containing flow, wherein the combustion reaction heats the supply material in the furnace; and during the operation of the furnace, the virtual sensor receives further condition data, and in response to the receipt of the further condition data, the virtual sensor transmits a signal to the furnace so that the virtual sensor uses the model to determine combustion products based on the further condition data and, based on the determined combustion products, automatically adjusts the flow rate of at least one of the first input section, the second input section, and the third input section.

[0030] Referring to Figure 1, a process 10 for operating the furnace according to an existing system is shown. Process 10 may include a step 12 in which at least one processor (e.g., a controller) determines whether physical sensors for measuring the combustion products of the furnace are present in and / or operational in the furnace. The combustion products are a composition of at least one compound that is generated and / or remains in the furnace after the combustion reaction, e.g., oxygen, CO, CO2, NO x This can refer to at least one of water and an unburned fuel composition. The physical sensor can be configured to directly measure combustion products in the furnace.

[0031] In step 14, the processor may determine that no physical sensors for measuring the combustion products of the furnace are present in the furnace and / or are not functioning. In existing systems, the absence of such physical sensors may result in a lack of combustion product information and potentially inefficient or suboptimal operation. This disclosure addresses such inefficiencies.

[0032] Continuing to refer to Figure 1, in step 16, the processor may determine that a physical sensor for measuring the combustion products of the furnace is present in and / or operating within the furnace. The physical sensor measures the combustion products (Y) within the furnace. p ) can be measured directly. In step 18, the processor can directly measure the desired combustion product (Y sp The processor can determine the desired combustion product, which may be automatically determined by the processor and / or a user-specified setpoint. In step 20, the processor can compare the measured combustion product with the desired combustion product to determine the deviation of the actual combustion product from the desired combustion product.

[0033] Continuing to refer to Figure 1, in step 22, the processor may decide whether to start a monitoring and / or control protocol. In step 24, the processor may start a monitoring protocol. The monitoring protocol may include the processor transmitting notifications and / or performing alarms (e.g., visual and / or auditory alarms) to inform the user of any deviation from the desired combustion products. In step 26, the processor may start a control protocol. The control protocol may include the processor automatically sending control signals to adjust at least one component of the furnace. For example, a control signal may cause a flow valve to adjust (e.g., open or close) the flow rate of at least one of the oxygen-containing flow, fuel-containing flow, and / or feed material flow into the furnace.

[0034] Referring to Figure 2, several embodiments or aspects of a process 30 for operating a furnace using a virtual gas sensor are shown. Process 30 may include a step 32 in which a processor determines whether a physical sensor for measuring the combustion products of the furnace is present in the furnace and / or operational. Unlike process 10 in Figure 1, the absence and / or deactivation of a physical sensor in the furnace for directly measuring the combustion products does not preclude the determination of the combustion products, as described herein.

[0035] Continuing to refer to Figure 2, in step 34, in response to the determination that the physical sensor is not present in the furnace and / or is not functioning, the processor can automatically acquire (e.g., receive and / or retrieve) further conditional data related to the operation of the furnace (e.g., by the combustion reaction). This further conditional data includes oxygen, CO, CO2, NO xThe composition does not include water and at least one of the unburned fuel. Thus, the further condition data does not include combustion products. The further condition data may include at least one of the total gas flow rate, fuel flow rate, oxygen flow rate, air flow rate, and furnace temperature that occur in the furnace during its operation. The further condition data may be an indirect indicator of combustion products, as determined by the model described herein. The further condition data may be measured by a physical sensor configured to measure the further condition data directly. The physical sensor configured to measure the further condition data may be different from a physical sensor configured to measure combustion products (which are not inputs to the model in this embodiment).

[0036] In step 36, the acquired further condition data can be input to a virtual sensor including a model as described herein. The virtual sensor uses the model to determine the combustion product (Y) based on the input further condition data. m ) can be determined. In step 38, the combustion products determined by the virtual sensor can be considered as the actual combustion products of the furnace.

[0037] In step 40, the processor generates the desired combustion product (Y sp ) can be determined. In step 42, the processor can compare the combustion product determined by the virtual sensor with the desired combustion product to determine the deviation of the actual combustion product from the desired combustion product.

[0038] In step 44, the processor may decide whether to start a monitoring and / or control protocol. In step 46, the processor may start a monitoring protocol. The monitoring protocol may include the processor transmitting notifications and / or performing alarms (e.g., visual and / or auditory alarms) to notify the user of any deviation from the desired combustion product. In step 48, the processor may start a control protocol. The control protocol may include the processor automatically sending control signals to adjust at least one component of the furnace. For example, a control signal may cause a flow valve to adjust (e.g., open or close) the flow rate of at least one of the oxygen-containing flow, fuel-containing flow, and / or feed material flow into the furnace.

[0039] Continuing to refer to Figure 2, in step 50, in response to the determination that a physical sensor is present in the furnace and / or operating in the furnace, the processor can automatically determine whether to calibrate the virtual sensor. If not, process 30 can proceed to step 34 as described above. If it is, process 30 can proceed to step 52.

[0040] In step 52, the processor may position the physical sensor to directly measure the combustion products in the furnace. Positioning the physical sensor may include activating the physical sensor, physically moving the physical sensor from a first position to a second position (e.g., stopping the retraction of the physical sensor), removing protective components from the physical sensor, or any combination thereof. Once positioned, in step 54, the physical sensor measures the combustion products in the furnace (Y p ) can be measured directly.

[0041] In step 56, the processor can automatically acquire (e.g., receive and / or retrieve) further condition data of the furnace. In step 58, the acquired further condition data can be input to a virtual sensor including a model as described herein. The virtual sensor uses the model to determine the combustion product (Y) based on the input further condition data. m ) can be determined. In step 60, the processor can determine further condition data, Y p and Y m Based on at least one of these, the virtual sensor (e.g., its model) can be recalibrated. During recalibration, Y measured directly by the physical sensor p And further condition data are considered to be the actual combustion products and further condition data, Y m This is considered to be the modeled combustion product, and the difference between the modeled combustion product and the actual combustion product is determined in step 60, and the model is recalibrated to more accurately predict the combustion product. This is the modeled combustion product Y m This could include updating the algorithm that generates it (for example, its parameters and / or the weights of those parameters).

[0042] Continuing to refer to Figure 2, in step 62, the processor determines whether calibration is complete. If yes, in step 66, the physical sensor may be positioned to protect it from the harsh conditions of the furnace. Positioning the physical sensor to protect it may include stopping the physical sensor, physically moving and / or removing the physical sensor from a second position to a first position (e.g., retracting or completely removing the physical sensor), installing protective components on the physical sensor, or any combination thereof.

[0043] If step 62 determines that calibration is not complete, step 64 determines the combustion product (Y) directly measured in the furnace. p) can be used as the actual combustion product, and process 30 can proceed to step 38 described above.

[0044] Referring to Figure 3, several non-limiting embodiments or aspects of a process 70 for generating a virtual gas sensor are shown. Process 70 may include a step 72 which includes determining the features of a model of the virtual sensor. The model features may include reaction chemistry data related to a combustion reaction (e.g., chemical reactions and / or their mass balance), inputs to the model (e.g., condition data), outputs to the model (e.g., combustion products), and model parameters (e.g., parameters that need to be fitted based on historical condition data). Model features such as inputs, parameters, and / or outputs may be determined automatically by the processor (e.g., using a machine learning algorithm) and / or specified or determined by the user.

[0045] In step 74, the processor may acquire (e.g., receive and / or retrieve) historical data. This historical data may include input and output data measured by physical sensors placed in the furnace during the training period, as further described herein. Physical sensors during the training period may directly measure combustion product and / or condition data of the furnace during the training period.

[0046] Continuing to refer to Figure 3, process 70 may include step 76, which generates a virtual sensor model based on acquired historical data. This may include fitting the model parameters based on the historical data to generate a model that accurately predicts combustion products based on the historical data. The analysis performed on the historical data to generate the model is not particularly limited and may include any type of analysis suitable for producing accurate results. Non-limiting examples of types of analysis that may be applied to historical data include nonlinear regression analysis, such as least-squares error nonlinear regression analysis. In step 78, the generated model is completed after a training period and can be implemented in the furnace.

[0047] Referring to Figure 4, a reactor system 80 using an existing system is shown. While this disclosure describes a specific design of reactor system 80 shown in Figures 4 to 6, it should be understood that this design is not limiting, and other reactor designs are also within the scope of this disclosure.

[0048] The furnace system 80 in Figure 4 may include a melting apparatus crown 82 into which a feed material is supplied and melted. The melting apparatus crown 82 may include a feed material input section 84 into which the feed material is supplied. The feed material may include any material intended to be melted by the furnace. An unrestricted example of a feed material is glass batch material used to produce glass substrates. An unrestricted list of glass batch materials includes at least one of the following: SiO2, Na2O, CaO, MgO, Al2O3, K2O, FeO, Fe2O3, SO3, Cr2O3, MnO, MnO2, TiO2, CeO2, B2O3, NiO, Li2O, CoO, V2O5, BaO, Cu, CuO, Co, Co3O4, Se, C, SnO2, ZrO2, Nd2O3, Er2O3, Sb2O3, As2O3, ZnO, NaNO3, etc. Other non-limiting examples of other raw materials that may be supplied to the melting apparatus crown 82 include reactants for chemical reactions, metallic materials (e.g., for smelting), and hydrocarbons for refining.

[0049] The melting apparatus crown 82 may include an oxygen-containing flow input section 86 through which an oxygen-containing flow is supplied to the melting apparatus crown 82. The oxygen-containing flow may include pure oxygen, air, or other oxygen-nitrogen gas blends.

[0050] The melting apparatus crown 82 may include a fuel-containing flow input section 88 into which a fuel-containing flow is supplied. The fuel may include any of the hydrocarbon fuels listed above.

[0051] Continuing to refer to Figure 4, each of the supply material input section 84, the oxygen-containing flow input section 86, and the fuel-containing flow input section 88 may be equipped with flow meters 90, 92, and 94, respectively, configured to directly measure the flow rates of the supply material, oxygen-containing flow, and fuel-containing flow to the melting apparatus crown 82. Each of the supply material input section 84, the oxygen-containing flow input section 86, and the fuel-containing flow input section 88 may also be equipped with valves (not shown) configured to control the amounts of the supply material, oxygen-containing flow, and fuel-containing flow supplied to the melting apparatus crown 82. The valves may be automatically controlled by a controller (not shown).

[0052] The melting apparatus crown 82 may also include temperature sensors 96a, 96b positioned within it to directly measure the temperature of specific areas within the furnace. A non-limiting example in Figure 4 shows two temperature sensors 96a, 96b, including a first temperature sensor 96a for measuring the temperature of a first area within the melting apparatus crown 82 and a second temperature sensor 96b for measuring the temperature of a second area within the melting apparatus crown 82. It will be understood that other embodiments may include only a single temperature sensor, and yet another embodiment may include three or more temperature sensors.

[0053] The melting apparatus crown 82 may further include a physical sensor 98 configured to directly measure combustion products within the melting apparatus crown 82. For example, the physical sensor 98 may be an oxygen sensor for directly measuring the oxygen composition within the melting apparatus crown 82. The combustion products may be any of the combustion products listed above. In Figure 4, the physical sensor 98 is located in the crown of the melting apparatus crown 82, but it will be understood that the physical sensor 98 may be positioned in one or more other areas of the melting apparatus crown 82 to directly measure combustion products.

[0054] Continuing to refer to Figure 4, the furnace system 80 may further include a port neck 100 equipped with a port neck physical sensor 102 configured to directly measure internal combustion products. The port neck 100 can connect the melting apparatus crown 82 to one or more accumulator crowns 104.

[0055] The furnace system 80 may further include a accumulator crown 104 equipped with a accumulator crown physical sensor 106 configured to directly measure internal combustion products. The accumulator crown 104 may include a target wall 108, which is equipped with a target wall physical sensor 110 configured to directly measure combustion products in the target wall 108.

[0056] Referring to Figures 5 and 6, furnace systems 80 having a virtual gas sensor 112 (interchangeably referred to as “virtual sensor 112”) are shown according to several aspects of the present disclosure. The furnace systems 80 of Figures 5 and 6 may include the same or similar components as the furnace system 80 of Figure 4, except as described below. The furnace systems 80 of Figures 5 and 6 may further comprise a virtual sensor 112 that communicates with a controller (not shown) of the furnace system 80. The virtual sensor 112 can replace and / or enhance at least one physical sensor of the furnace system 80 configured to directly measure combustion products. A non-limiting example of Figure 5 or 6 shows a virtual sensor 112 replacing and / or enhancing a physical sensor 98 of the melting apparatus crown 82, but it will be understood that the virtual sensor 112 can additionally or alternatively replace and / or enhance at least one physical sensor 102, 106, 110 of the port neck 100, accumulator crown 104, and target wall 108.

[0057] Referring to Figure 5, the furnace system 80 may be equipped with a virtual sensor 112, and the physical sensor 98 (see Figure 4) is removed from the melting apparatus crown 82. The physical sensor 98 can be completely removed from the melting apparatus crown 82 to protect it from the harsh conditions within the melting apparatus crown 82 and thus extend the life of the physical sensor 98. In some non-limiting embodiments, the physical sensor 98 can be periodically reinserted into the melting apparatus crown 82 to evaluate the deviation of the virtual sensor 112 and / or to recalibrate the virtual sensor 112.

[0058] Referring to Figure 6, the furnace system 80 may be equipped with a virtual sensor 112, and the physical sensor 98 is retracted from the melting apparatus crown 82 (compared to Figure 4). Retracting the physical sensor 98 from the melting apparatus crown 82 protects the physical sensor 98 from the harsh conditions within the melting apparatus crown 82 and thus extends the life of the physical sensor 98. In some non-limiting embodiments, the physical sensor 98 may periodically cease retracting from the melting apparatus crown 82 in order to evaluate the deviation of the virtual sensor 112 and / or to recalibrate the virtual sensor 112.

[0059] Continuing with reference to Figures 5 and 6, the physical sensor 98 is shown completely removed (Figure 5) or retracted (Figure 6) in conjunction with the virtual sensor 112, which is configured to indirectly measure combustion products based on a model, but other arrangements of the physical sensor 98 are also within the scope of this disclosure. For example, the physical sensor 98 may remain in the position shown in Figure 4, but simply stopped for the virtual sensor 112. In some non-limiting embodiments, the physical sensor 98 and the virtual sensor 112 can be used to simultaneously measure or model combustion products, for example, during model recalibration. In some non-limiting embodiments, the physical sensor 98 may remain in the position shown in Figure 4, but the physical sensor 98 may be covered by a separate component to protect the physical sensor from harsh conditions within the melting apparatus crown 82.

[0060] Referring to Figure 7, a model training system 120 according to several aspects of the present disclosure is shown, which can be used to generate a virtual sensor 112 for use with a furnace system 80 from Figures 5 and 6. The model training system 120 may comprise a model building system 122, which comprises at least one processor configured to generate a virtual sensor 112 containing a model based on input to the model building system 122. The virtual sensor 112 generated by the model building system 122 may be configured to determine combustion products in the furnace.

[0061] Continuing to refer to Figures 5 to 7, the input to the model building system 122 may include condition data, including the operating conditions of the furnace, collected by physical sensors 98 pre-placed within the furnace (e.g., the melting unit crown 82; in this specification, “furnace” and “melting unit crown” may be interchangeably referred to). The operating conditions of the furnace 82 can be collected by physical sensors 98 pre-placed within the furnace 82 during the training period, for example, by measuring combustion products during the training period. The physical sensors 98 can be placed during the training period in the crown of the furnace 82, the accumulator crown 104, or any other suitable location in the furnace system 80.

[0062] During the training period (before the provision of the virtual sensor 112), the physical sensor 98 can be placed inside the furnace 82, which can be operated by the combustion reaction of an oxygen-containing flow that burns a fuel-containing flow to heat and / or melt the feed material. During the training period, the physical sensor 98 can collect conditional data of the furnace 82, such as by measuring the combustion products inside the furnace 82 during the training period. During the training period, other physical sensors in the furnace system, such as meters 90, 92, 94, temperature sensors 96a, 96b, or other physical sensors (e.g., 102, 106, 110), can measure conditional data of the furnace system 80. Non-limiting examples of conditional data that may be measured during the training period include at least one of fuel flow rate, oxygen flow rate, air flow rate, furnace temperature, oxygen composition, carbon dioxide composition, and carbon monoxide composition.

[0063] Conditional data measured during this training period can be considered historical data, and this historical data can be input into the model building system 122. Reaction chemistry data 124 related to the combustion reaction can also be input into the model building system 122. Reaction chemistry data 124 may include data related to the execution of the combustion reaction, such as the chemical formula used during the combustion reaction, including the amount (e.g., moles) of each reactant used and / or the product formed. Reaction chemistry data 124 may also include the mass balance of the combustion reaction.

[0064] Based on the aforementioned inputs, the model building system 122 can generate a model that includes an algorithm for accurately predicting combustion products based on historical input data. This may include determining inputs and / or fitting parameters for accurately predicting combustion products. The model building system 122 may be equipped with machine learning algorithms and / or use any other suitable fitting program. The output of the model building system 122 may include a virtual sensor 112 containing the model.

[0065] Referring to Figure 8, a model implementation system 125 according to several aspects of this disclosure is shown, which can be used as a component of the furnace system 80 from Figures 5 and 6. Integration of the implementation system 125 with the furnace system 80 can be achieved by providing the furnace system 80 described herein and integrating the virtual sensor 112 as a component of the furnace system 80 by providing the generated virtual sensor 112 to the furnace system 80.

[0066] Referring to Figures 5, 6, and 8, the virtual sensor 112 may be configured to communicate with the reactor system 80, for example, its controller 128. The controller 128 can communicate with, for example, a physical sensor 98 (when operating), flow meters 90, 92, 94, temperature sensors 96a, 96b, other physical sensors (e.g., 102, 106, 110), a feed material input 84, an oxygen-containing flow input 86, and valves (not shown) in the fuel-containing flow input 88.

[0067] Continuing to refer to Figures 5, 6, and 8, the furnace system 80 is operated by a combustion reaction (e.g., in the melting unit crown 82) generated by combining oxygen from an oxygen-containing flow and fuel from a fuel-containing flow, which can heat the feed material in the melting unit crown 82. Heating the feed material can include melting the feed material. During this operation of the furnace (after the training period, when the virtual sensor 112 is integrated and operating), the virtual sensor 112 can receive further conditional data (as previously defined). For example, the virtual sensor 112 can receive measurement data during furnace operation from flow meters 90, 92, 94, temperature sensors 96a, 96b, other physical sensors (e.g., 102, 106, 110), valves (not shown) in the feed material input 84, oxygen-containing flow input 86, and fuel-containing flow input 88, etc. The virtual sensor 112 can also receive reaction chemistry data 124 related to the combustion reaction of the furnace system 80. Therefore, at some point during the operation of the reactor, the reactor system 80 may be equipped with at least one of the fuel flow sensor 94, oxygen or air flow sensor 92, and temperature sensors 96a, 96b, and the reactor system 82 does not use the readings of the physical sensor 98 as input for determining the combustion products 126 in the reactor system 80.

[0068] In response to the reception of this further conditional data, the virtual sensor 112 can use a model to automatically determine the combustion products 126 based on the input further conditional data. In some non-limiting embodiments or aspects, the virtual sensor 112 determines the combustion products 126 without input from the physical sensor 98, and the readings from the physical sensor 98 are not input for determining the combustion products 126 in the furnace 82 at some point during the operation of the furnace system 80. The combustion products 126 are O2, CO, CO2, NO2, etc., determined by the virtual sensor 112 based on the further conditional data. x This may include determining the composition of at least one of the following: water and unburned fuel.

[0069] Based on the determined combustion product 126, the virtual sensor 112 can send a signal to the controller 128 of the furnace system 80 to automatically adjust at least one setting of the furnace system. For example, the controller 128 can automatically adjust the flow rate of at least one of the supply material input 84, oxygen-containing inlet 86, and fuel-containing inlet 88, such as by further opening or closing a valve. For example, the controller 128 can automatically adjust the flow rate of the oxygen-containing inlet 86 to allow a higher or lower flow rate of oxygen to reach the furnace system 80.

[0070] The controller 128 can adjust at least one of the supply material input section 84, the oxygen-containing flow input section 86, and the fuel-containing flow input section 88 based on at least one predetermined threshold level of combustion products. For example, in response to the virtual sensor 112 determining that the combustion products are below a threshold level, the controller 128 can adjust at least one of the supply material input section 84, the oxygen-containing flow input section 86, and the fuel-containing flow input section 88 by performing a first action, and in response to the virtual sensor determining that the combustion products are above a threshold level, the controller 128 can adjust at least one of the supply material input section 84, the oxygen-containing flow input section 86, and the fuel-containing flow input section 88 by performing a second different action. These adjustments can be performed automatically by the controller 128. Additionally or alternatively, the controller 128 can adjust the furnace burner based on at least one predetermined threshold level of combustion products.

[0071] For example, in response to the oxygen level in the combustion products exceeding a predetermined threshold, the controller 128 can decrease the oxygen-containing inlet 86 and / or increase the fuel-containing inlet 88 and / or adjust the furnace burner. For example, in response to the oxygen level in the combustion products falling below a predetermined threshold, the controller 128 can increase the oxygen-containing inlet 86 and / or decrease the fuel-containing inlet 88 and / or adjust the furnace burner. For example, in response to the CO level in the combustion products exceeding a predetermined threshold, the controller 128 can increase the oxygen-containing inlet 86 and / or decrease the fuel-containing inlet 88 and / or adjust the furnace burner. For example, NO in the combustion products x Based on the level, the controller 128 can adjust the oxygen-containing inlet 86 and / or the fuel-containing inlet 88, and / or the furnace burner. For example, based on the CO:CO2 ratio in the combustion products, the controller 128 can adjust the oxygen-containing inlet 86 and / or the fuel-containing inlet 88, and / or the furnace burner.

[0072] The furnace system 80 can be operated continuously using a virtual sensor 112 that continuously determines the combustion products 126, and the virtual sensor 112 can communicate signals to the controller 128 to appropriately adjust at least one characteristic of the furnace during this continuous operation of the furnace system 80.

[0073] Referring to Figure 9, several embodiments of the model recalibration system 130 of this disclosure are shown, which can be used as a component of the reactor system 80 in Figures 5 and 6. The model recalibration system 130 can be used to recalibrate (e.g., update) a previously generated virtual sensor 112.

[0074] Referring to Figures 5, 6, and 9, the physical sensor 98 can be activated and / or reintroduced into the reactor system 80 (e.g., inserted, de-evacuated, uncovered, etc.) to recalibrate the virtual sensor 112. The physical sensor 98 and the virtual sensor 112 can communicate with the model building system 122. Flow meters 90, 92, 94, temperature sensors 96a, 96b, other physical sensors (e.g., 102, 106, 110), the feed material input 84, the oxygen-containing flow input 86, and valves (not shown) in the fuel-containing flow input 88 can communicate with the model building system 122. Reaction chemistry data 124 can communicate with the model building system 122.

[0075] During furnace operation to recalibrate the virtual sensor 112, the physical sensor 98 can be activated while the virtual sensor 112 receives further conditional data and determines the combustion products. The physical sensor 98 can determine the measured combustion products by directly measuring them. The combustion products measured by the physical sensor 98 can be considered as the actual combustion products, while the combustion products determined by the virtual sensor 112 can be considered as the modeled combustion products. The deviation of the virtual sensor 112 can be determined by comparing the measured combustion products with the modeled combustion products. It will be understood that the measured and modeled combustion products correspond to the simultaneously measured and modeled combustion products during furnace operation, so that the relevant combustion products can be compared.

[0076] In response to the determination of the deviation, the model building system 122 can determine whether the deviation exceeds a threshold. In response to the deviation exceeding the threshold, the model building system 122 can update the model of the virtual sensor 112 based on the measured combustion products (measured by the physical sensor 98). The model can also be updated based on other inputs to the model building system 122, such as at least one of the flow meters 90, 92, 94, temperature sensors 96a, 96b, other physical sensors (e.g., 102, 106, 110), the valves (not shown) of the feed material input section 84, the oxygen-containing flow input section 86, and the fuel-containing flow input section 88, and reaction chemistry data 124. The model building system 122 can generate an updated virtual sensor 112u containing the updated model. The updated model may have different (e.g., updated) parameters than the previous model. The updated virtual sensor 112u is integrated into the reactor system 80 and can be used to determine the combustion products when the physical sensor 98 is stopped again, and / or has been evacuated from and / or removed from the reactor system 80.

[0077] To keep the virtual sensor 112 accurate, the virtual sensor 112 can be periodically recalibrated using the model recalibration system 130 described herein. The virtual sensor 112 can be recalibrated automatically and / or in response to a user performing a function to request recalibration of the virtual sensor 112.

[0078] Although the embodiments described above describe a furnace system 80 having a single virtual sensor, it will be understood that the furnace system 80 may have multiple virtual sensors, such as a first virtual sensor having a first model and a second virtual sensor having a second different model, and may also include other virtual sensors. Each virtual sensor can be generated as described above.

[0079] In some non-limiting embodiments or aspects, a first model can be generated using inputs that include condition data relating to the operating conditions of a first location within the furnace system, and a second model can be generated using inputs that include condition data relating to the operating conditions of a second location within the furnace system. For example, the first location within the furnace system may include the smelter crown, and the second location may include at least one of the port neck, accumulator crown, target wall, etc. For example, the first location within the furnace system may include a first location of the smelter crown, and the second location may include a second location of the smelter crown (e.g., two different locations within the same furnace component). The use of multiple virtual sensors can account for variable conditions present in different regions of the furnace and enable more accurate modeling of combustion products by considering the furnace system conditions that have the most influence when generating accurate models of combustion products in different regions of the furnace. Different conditions within the furnace may result in one region being closer to or further from the location of the furnace burner where the furnace is burning, such as the right side of the furnace having a higher temperature than the left side of the furnace when the furnace is burning from the right side. When generating different models, different condition data within the reactor system can be used (for example, as input for generating the models) to generate different models.

[0080] In some non-limiting embodiments, a single physical sensor located within the furnace system may replace and / or enhance a single virtual sensor. However, in some non-limiting embodiments, a single physical sensor located within the furnace system may replace and / or enhance multiple virtual sensors, for example, when two different models improve the accuracy of determining combustion products pre-measured by a single physical sensor.

[0081] The disclosure also relates to a virtual sensor comprising at least one processor that stores a model for determining combustion products in a furnace. The model can be generated as described above using inputs that include condition data, including furnace operating conditions, collected by physical gas sensors pre-placed in the furnace.

[0082] The disclosure also relates to a plurality of virtual sensors, including a first virtual sensor including a first model and a second virtual sensor including a second model, wherein the first model is generated using an input that includes condition data relating to the operating conditions of a first location in the furnace, and the second model is generated using an input that includes condition data relating to the operating conditions of a second location in the furnace.

[0083] The following numbered sections illustrate various aspects of this disclosure.

[0084] Item 1: A method for operating a furnace using a virtual gas sensor, comprising: providing a furnace comprising a first input containing a feed material configured to be heated in the furnace; a second input containing an oxygen-containing flow; and a third input containing a fuel-containing flow; providing a virtual sensor including a model for determining combustion products in the furnace, wherein the model is generated using an input containing condition data, the condition data including operating conditions of the furnace collected by a physical gas sensor pre-placed in the furnace; installing the virtual sensor to communicate with the furnace so that it is configured to receive further condition data relating to the operating conditions of the furnace; and operating the furnace by a combustion reaction produced by combining oxygen from the oxygen-containing flow and fuel from the fuel-containing flow, wherein the combustion reaction heats the feed material in the furnace; and during the operation of the furnace, the virtual sensor receives further condition data, and in response to the receipt of the further condition data, the virtual sensor transmits a signal to the furnace so that the virtual sensor uses the model to determine combustion products based on the further condition data and, based on the determined combustion products, automatically adjusts the flow rate of at least one of the first input, the second input, and the third input.

[0085] Item 2: The method according to Item 1, further comprising the steps of: placing a physical gas sensor inside a furnace; operating the furnace by a combustion reaction during a training period prior to providing a virtual sensor; collecting conditional data using the physical gas sensor during the training period; and generating a model using the conditional data collected during the training period.

[0086] Item 3: Combustion products are oxygen, CO, CO2, NO x The method according to claim 1 or 2, comprising at least one of the following: and an unburned fuel composition.

[0087] Item 4: The method according to any one of items 1 to 3, wherein the supplied material includes glass batch material.

[0088] Item 5: The method according to any of Items 1-4, wherein physical sensors are placed in the furnace crown and / or accumulator to collect conditional data, and the readings of the physical gas sensors are not inputs for determining the combustion products in the furnace at any point in time during furnace operation.

[0089] Item 6: The method according to any one of Items 1 to 5, wherein the condition data includes at least one of fuel flow rate, oxygen flow rate, air flow rate, furnace temperature, oxygen composition, carbon dioxide composition, and carbon monoxide composition.

[0090] Item 7: Further condition data, oxygen, CO, CO2, NO x The virtual sensor does not contain at least one composition of the unburned fuel, and based on further conditional data, it detects oxygen, CO, CO2, NO x The method according to claim 6, and determining the composition of at least one of the unburned fuels.

[0091] Item 8: A method according to any of Items 1-7, wherein a model is generated based on conditional data and reaction chemistry.

[0092] Item 9: The method according to any one of items 1 to 8, wherein a signal transmitted to the furnace automatically adjusts the flow rate of oxygen-containing flow into the furnace.

[0093] Item 10: The method according to any one of items 1 to 9, further comprising the steps of: activating physical sensors in the furnace while a virtual sensor is receiving further conditional data during furnace operation; determining measured combustion products in the furnace using the physical sensors; and comparing the measured combustion products with the combustion products determined by the virtual sensor to determine the difference.

[0094] Item 11: The method according to item 10, further comprising the step of updating the model using the measured combustion product in response to the deviation exceeding a threshold.

[0095] Item 12: The method according to any one of Items 1 to 11, wherein at some point during the operation of the furnace, the furnace is equipped with at least one of a fuel flow sensor, an oxygen flow sensor, an air flow sensor, and a temperature sensor, and the furnace does not use readings from a physical gas sensor as input for determining the combustion products in the furnace.

[0096] Item 13: The method according to any one of Items 1 to 12, comprising the step of providing a plurality of virtual sensors, including a first virtual sensor including a first model and a second virtual sensor including a second model, wherein the first model is generated using an input including condition data relating to operating conditions at a first location in a furnace, and the second model is generated using an input including condition data relating to operating conditions at a second location in a furnace.

[0097] Item 14: A furnace system comprising: a first input containing a supply material configured to be heated within the furnace; a second input containing an oxygen-containing flow; and a third input containing a fuel-containing flow, and configured to operate by a combustion reaction produced by combining oxygen from the oxygen-containing flow and fuel from the fuel-containing flow; and a virtual sensor including a model for determining combustion products within the furnace, the model being generated using an input containing condition data, the condition data including operating conditions of the furnace collected by a physical gas sensor pre-placed within the furnace, the virtual sensor communicating with the furnace such that the virtual sensor is configured to receive further condition data relating to the operating conditions of the furnace, the virtual sensor being configured to determine the combustion products of the furnace in response to the receipt of further condition data, and the virtual sensor being configured to transmit a signal to the furnace to automatically adjust the flow rate of at least one of the first input, the second input, and the third input based on the determined combustion products.

[0098] Item 15: The system according to Item 14, further comprising at least one of a fuel flow sensor, an oxygen flow sensor, an air flow sensor, and a temperature sensor located inside the reactor.

[0099] Item 16: The system described in Item 14 or 15, wherein at some point during furnace operation, the furnace system does not use readings from a physical gas sensor as input for determining combustion products in the furnace.

[0100] Item 17: Combustion products in the furnace are oxygen, CO, CO2, NO x The system according to any one of claims 14 to 16, comprising at least one of the following: and an unburned fuel composition.

[0101] Item 18: A system according to any one of items 14 to 17, wherein the supplied material includes glass batch material.

[0102] Item 19: A system according to any one of items 14 to 18, wherein the condition data includes at least one of the following: fuel flow rate, oxygen flow rate, air flow rate, furnace temperature, oxygen composition, carbon dioxide composition, and carbon monoxide composition.

[0103] Item 20: Further condition data, oxygen, CO, CO2, NO x The virtual sensor does not contain at least one composition of the unburned fuel, and based on further conditional data, it detects oxygen, CO, CO2, NO x The system according to paragraph 19, which determines the composition of at least one of the unburned fuels.

[0104] Item 21: A system described in any of items 14-20, in which the model is generated based on conditional data and reaction chemistry.

[0105] Item 22: The system according to any one of items 14 to 21, wherein a signal transmitted to the furnace is configured to automatically adjust the flow rate of oxygen-containing flow into the furnace.

[0106] Item 23: A system according to any one of items 14-22, wherein the model is configured to be updated by activating physical sensors in the furnace while virtual sensors receive further conditional data, using the physical sensors to determine the measured combustion products in the furnace, comparing the measured combustion products with the combustion products determined by the virtual sensors to determine the deviation, and updating the model with the measured combustion products in response to the deviation exceeding a threshold.

[0107] Item 24: A system according to any one of items 14 to 23, comprising providing a plurality of virtual sensors including a first virtual sensor including a first model and a second virtual sensor including a second model, wherein the first model is generated using an input including condition data relating to the operating conditions of a first location in the furnace, and the second model is generated using an input including condition data relating to the operating conditions of a second location in the furnace.

[0108] Item 25: A virtual sensor comprising at least one processor for storing a model for determining combustion products in a furnace, wherein the model is generated using inputs including conditional data, the conditional data including furnace operating conditions collected by a physical gas sensor pre-placed in the furnace.

[0109] Item 26: The virtual sensor according to Item 25, wherein the furnace comprises a first input section containing a feed material configured to be heated within the furnace, a second input section containing an oxygen-containing flow, and a third input section containing a fuel-containing flow, and the furnace is configured to operate by a combustion reaction produced by combining oxygen from the oxygen-containing flow and fuel from the fuel-containing flow.

[0110] Item 27: The virtual sensor according to Item 26, wherein at least one processor is programmed or configured to communicate with the furnace to receive further condition data relating to the operating conditions of the furnace, to determine the combustion products of the furnace in response to the receipt of further condition data, and to send a signal to the furnace to automatically adjust the flow rate of at least one of the first input, second input, and third input based on the determined combustion products.

[0111] Item 28: A virtual sensor as described in any of items 25-27, wherein at some point during furnace operation, the furnace system does not use readings from a physical gas sensor as input for determining combustion products within the furnace.

[0112] Item 29: Combustion products in the furnace are oxygen, CO, CO2, NO x A virtual sensor according to any one of claims 25 to 28, comprising at least one of the following: and an unburned fuel composition.

[0113] Item 30: A virtual sensor according to any of items 25 to 29, wherein the supplied material includes glass batch material.

[0114] Item 31: A virtual sensor as described in any of items 25 to 30, wherein the condition data includes at least one of fuel flow rate, oxygen flow rate, air flow rate, furnace temperature, oxygen composition, carbon dioxide composition, and carbon monoxide composition.

[0115] Item 32: Further condition data, oxygen, CO, CO2, NO x , and does not contain at least one composition of unburned fuel, and at least one processor, based on further condition data, oxygen, CO, CO2, NO x A virtual sensor according to any one of paragraphs 27 to 31, which determines the composition of at least one of the unburned fuels.

[0116] Item 33: A virtual sensor described in any of items 25-32, the model of which is generated based on conditional data and reaction chemistry.

[0117] Item 34: A virtual sensor as described in any of items 27-33, wherein the signal transmitted to the furnace is configured to automatically adjust the flow rate of oxygen-containing flow into the furnace.

[0118] Item 35: A virtual sensor according to any one of items 27 to 34, wherein at least one processor is further programmed or configured to receive measured combustion products in the furnace, determined by physical sensors in the furnace, while the virtual sensor is receiving further condition data, to compare the measured combustion products with the combustion products determined by the virtual sensor to determine a deviation, and to update the model based on the measured gas content in response to the deviation being above a threshold.

[0119] Item 36: A plurality of virtual sensors as described in any of items 25 to 35, comprising a first virtual sensor including a first model and a second virtual sensor including a second model, wherein the first model is generated using an input including condition data relating to the operating conditions of a first location in the furnace, and the second model is generated using an input including condition data relating to the operating conditions of a second location in the furnace.

[0120] Examples Example 1 Generating a model for a virtual sensor Referring to Figure 10, non-limiting examples of modeling processes 140 according to some aspects of the present disclosure are shown. Modeling processes 140 can be used to generate models of virtual sensors. Modeling processes 140 may include determining reaction chemistry data 124 of a combustion reaction, such as the reactions that occur in the combustion reaction and their mass balance.

[0121] The modeling process 140 may include determining input variables 142 that are input to the modeling construction system to generate the model. The modeling process 140 may also include determining output variables 144 (e.g., combustion products) that are desired to be determined by the model. The modeling process 140 may also include determining parameters 146 that enable the accurate generation of combustion products, which may include determining the most relevant parameters and / or the weighting of those parameters. The modeling process 140 may output a model, which may include an algorithm 148 for determining the combustion products when integrated into a furnace system, based on the measured input parameters.

[0122] Example 2 Testing of a melting apparatus crown with a single virtual sensor Referring to Figure 11, a virtual sensor was generated for the melting apparatus crown (crown #2) in accordance with this disclosure, and the results from the generated virtual sensor were tested against a physical oxygen (e.g., combustion products) sensor located on the melting apparatus crown. A single virtual sensor was generated to represent a single physical oxygen sensor, or the measurements from a single physical sensor were averaged over different operating conditions. Test data 150 showing the oxygen composition measured from the physical oxygen sensor was plotted over several weeks against the modeled oxygen composition determined by the virtual sensor. As can be seen from the test data 150, the oxygen composition modeled by the virtual sensor closely matches the oxygen composition measured simultaneously by the physical oxygen sensor, indicating that a high-precision virtual sensor representing the readings of the physical oxygen sensor was generated.

[0123] Example 3 Testing of a melting apparatus crown with multiple virtual sensors Referring to Figures 12A and 12B, multiple virtual sensors were generated for the melting apparatus crown (crown #2) in accordance with this disclosure, and the results from the generated virtual sensors were tested against a physical oxygen (e.g., combustion product) sensor located in the melting apparatus crown. One virtual sensor was generated to represent a physical oxygen sensor located on the left side of the furnace or a physical oxygen sensor burning from the left side, and different virtual sensors were generated to represent a physical oxygen sensor located on the right side or a physical oxygen sensor burning from the right side of the furnace.

[0124] Test data 150L shows the oxygen composition measured from the left-side physical oxygen sensor plotted over several weeks, compared to the modeled oxygen composition determined by the left-side virtual sensor. As can be seen from test data 150L, ​​the oxygen composition modeled by the left-side virtual sensor closely matches the oxygen composition measured by the left-side physical oxygen sensor over the same period, indicating that a highly accurate left-side virtual sensor representing the readings of the left-side physical oxygen sensor has been generated.

[0125] Test data 150R shows the oxygen composition measured from the right-side physical oxygen sensor plotted over several weeks, compared to the modeled oxygen composition determined by the right-side virtual sensor. As can be seen from test data 150R, the oxygen composition modeled by the right-side virtual sensor closely matches the oxygen composition measured by the right-side physical oxygen sensor over the same period, indicating that a highly accurate right-side virtual sensor representing the readings of the right-side physical oxygen sensor has been generated.

[0126] Those skilled in the art will readily understand that modifications to the present invention can be made without departing from the concepts disclosed in the foregoing description. Accordingly, the specific embodiments described in detail herein are illustrative and do not limit the scope of the present invention, which should be given the entire scope of the appended claims and all equivalents.

Claims

1. A method for operating a reactor using a virtual gas sensor, A step of providing a furnace comprising a first input section containing a supply material configured to be heated in the furnace, a second input section containing an oxygen-containing flow, and a third input section containing a fuel-containing flow. A step of providing a virtual sensor including a model for determining combustion products in the furnace, wherein the model is generated using inputs including condition data, the condition data including operating conditions of the furnace collected by a physical gas sensor pre-placed in the furnace; The steps include: installing the virtual sensor to communicate with the furnace so that it is configured to receive further condition data related to the operating conditions of the furnace; A step of operating the furnace by a combustion reaction generated by combining oxygen from the oxygen-containing stream and fuel from the fuel-containing stream, wherein the combustion reaction heats the supply material in the furnace, and during the operation of the furnace, The virtual sensor receives the further condition data, In response to receiving the aforementioned further conditional data, the virtual sensor uses the model to determine the combustion products based on the aforementioned further conditional data, and Based on the determined combustion product, the virtual sensor transmits a signal to the furnace in order to automatically adjust the flow rate of at least one of the first input unit, the second input unit, and the third input unit. Steps and Methods that include...

2. The steps include placing the physical gas sensor inside the furnace, The steps include operating the furnace by the combustion reaction during a training period prior to providing the virtual sensor, During the aforementioned training period, the steps include: collecting the condition data using the physical gas sensor; A step of generating the model using the conditional data collected during the training period. The method according to claim 1, further comprising:

3. The combustion products are oxygen, CO, CO ２ NO ｘ The method according to claim 1 or 2, comprising at least one of the following: and an unburned fuel composition.

4. The method according to any one of claims 1 to 3, wherein the supply material includes a glass batch material.

5. The method according to any one of claims 1 to 4, wherein the physical sensor is positioned in the crown and / or accumulator of the furnace to collect the condition data, and the reading of the physical gas sensor is not an input for determining the combustion products in the furnace at any point in time during the operation of the furnace.

6. The method according to any one of claims 1 to 5, wherein the condition data includes at least one of fuel flow rate, oxygen flow rate, air flow rate, furnace temperature, oxygen composition, carbon dioxide composition, and carbon monoxide composition.

7. The aforementioned further condition data is oxygen, CO, CO ２ NO ｘ , and does not contain at least one composition of unburned fuel, and the virtual sensor, based on the further condition data, contains oxygen, CO, CO ２ NO ｘ The method according to claim 6, which determines the composition of at least one of the unburned fuels.

8. The method according to any one of claims 1 to 7, wherein the model is generated based on the condition data and reaction chemistry.

9. The method according to any one of claims 1 to 8, wherein the signal transmitted to the furnace automatically adjusts the flow rate of the oxygen-containing flow to the furnace.

10. During the operation of the furnace, the steps include activating the physical sensors in the furnace while the virtual sensors are receiving the further condition data, The steps include determining the measured combustion products in the furnace using the aforementioned physical sensors, The steps include comparing the measured combustion product with the combustion product determined by the virtual sensor to determine the difference, and The method according to any one of claims 1 to 9, further comprising:

11. In response to the deviation exceeding a threshold, the model is updated using the measured combustion products. The method according to claim 10, further comprising:

12. The method according to any one of claims 1 to 11, wherein at some point during the operation of the furnace, the furnace is equipped with at least one of a fuel flow sensor, an oxygen flow sensor, an air flow sensor, and a temperature sensor, and the furnace does not use the readings of the physical gas sensor as input for determining the combustion products in the furnace.

13. A step of providing a plurality of virtual sensors, including a first virtual sensor including a first model and a second virtual sensor including a second model, The first model is generated using an input that includes condition data relating to the operating conditions of a first location within the furnace, and the second model is generated using an input that includes condition data relating to the operating conditions of a second location within the furnace. Step The method according to any one of claims 1 to 12, including the method described above.

14. A furnace comprising a first input section containing a supply material configured to be heated within the furnace, a second input section containing an oxygen-containing flow, and a third input section containing a fuel-containing flow, and configured to operate by a combustion reaction generated by combining oxygen from the oxygen-containing flow and fuel from the fuel-containing flow, A virtual sensor including a model for determining combustion products in the furnace, wherein the model is generated using inputs including condition data, and the condition data includes the operating conditions of the furnace collected by physical gas sensors pre-placed in the furnace. Equipped with, The virtual sensor communicates with the furnace so that it is configured to receive further condition data related to the operating conditions of the furnace. The virtual sensor is configured to determine the combustion products of the furnace in response to the reception of the further condition data. The virtual sensor is configured to transmit a signal to the furnace to automatically adjust the flow rate of at least one of the first input unit, the second input unit, and the third input unit based on the determined combustion product. Furnace system.

15. The system according to claim 14, further comprising at least one of a fuel flow sensor, an oxygen flow sensor, an air flow sensor, and a temperature sensor disposed within the furnace.

16. The system according to claim 14 or 15, wherein at some point during the operation of the furnace, the furnace system does not use the readings of the physical gas sensor as input for determining the combustion products in the furnace.

17. The combustion products in the furnace are oxygen, CO, CO ２ NO ｘ The system according to any one of claims 14 to 16, comprising at least one of the following: and an unburned fuel composition.

18. The system according to any one of claims 14 to 17, wherein the supply material includes a glass batch material.

19. A virtual sensor comprising at least one processor for storing a model for determining combustion products in a furnace, wherein the model is generated using inputs including conditional data, and the conditional data includes operating conditions of the furnace collected by physical gas sensors pre-placed in the furnace.

20. A plurality of virtual sensors according to claim 19, comprising a first virtual sensor including a first model and a second virtual sensor including a second model, The first model is generated using an input that includes condition data relating to the operating conditions of a first location within the furnace, and the second model is generated using an input that includes condition data relating to the operating conditions of a second location within the furnace. Multiple virtual sensors.

Images