Control of peeling system

Abstract

A method of controlling a peeling system is provided, including monitoring at least one sensor configured to detect data indicative of a characteristic of a flow of peelable product through the system; providing the data to a virtual model of the peeling system; the operation of the peeling system is adjusted based on the output of the virtual model.

Description

Technical Field

[0001] The present invention relates to a peeling system and, in particular, to controlling a peeling system using data provided from a virtual representation of the system. Background Art

[0002] Mechanical peeling methods are used in many food processing lines, including potato processing lines for striping and slicing potato products. Peeling machines can take many forms, including steam peelers, grinding peelers, and knife peelers. Other forms of peeling machines can use chemical peeling methods, such as caustic and lye peeling. In some cases, peeling equipment will also be used to remove some surface defects and flaws. The term "peeling" as used herein is intended to refer to any form of external surface removal, including but not limited to peeling by steam, grinding, knife, chemical or polishing.

[0003] Peeling systems remove a thin layer of the product being peeled, ideally only that layer containing the husk or shell, with minimal or no removal and loss of the adjacent quality product flesh. However, some quality product is removed, and all peeling systems must deal with the conflict of removing the most husk and surface defects while minimizing the loss of quality product. The greatest challenge facing peeling systems is to minimize the loss of quality material while maximizing the removal of husk or unwanted outer surfaces. Within the scope of the present invention, the term "peeling aggressiveness" is used to refer to this balance to be achieved between over-peeling and under-peeling. A secondary but important issue for industries using peeling systems is the energy consumption of the system, which should be minimized while still removing the husk at an economical productivity rate. Another issue is the large amount of waste and contaminated water generated by the peeling process, which needs to be filtered, treated and disposed of.

[0004] Peeling quality depends on many factors, including, for example, product size and quality (including product defects) and pressure and temperature at various points in the peeling system. A change in one of these factors during the peeling process, or indeed, a miscalculation or misunderstanding of one of these factors before the peeling process begins, can adversely affect peeling quality. Furthermore, the peeling process may have been in progress for some time before deterioration in peeling quality is detected. This is generally because such deterioration is usually apparent by means of visual inspection by personnel on site at the system or by visual inspection of a camera trained on the output of the system.

[0005] Furthermore, many of the inefficiencies in current peeling systems are caused by downtime in the peeling process. Downtime is often caused by stopping the process in order to change a parameter in the system. The trigger for such a parameter change is often an indication of deterioration in the peel quality of the system output, which is also often noticed by visual inspection. Stopping and starting the process in this manner is inefficient and further wastes energy. An example is the energy used to return a piece of equipment to optimal operating temperature. These issues result in a combination of inefficiencies in the stop / start process and variations in product output quality.

[0006] A control system that takes into account real-time data from the peeling process to ensure an optimized peeling process and desired peeled product would be an improvement over the prior art. This would provide an energy efficient and cost effective peeling system that improves peeling quality and is more reliable with less system downtime. Summary of the invention

[0007] A method of controlling a peeling system is provided, comprising:

[0008] Monitor the peeling process of the peeling system and the products peeled by the peeling system in real time using:

[0009] at least a first sensor for monitoring at least one characteristic of the product, and

[0010] Used to monitor process operations;

[0011] Obtaining data from at least a first sensor to perform a simulation, the simulation simulating a peeling process and a product flow through a peeling system in a virtual model, the simulation comprising;

[0012] comparing at least one simulated output of the virtual model to at least one desired output of the peeling system using an optimizer of the virtual model;

[0013] using the output of the optimizer to adjust control parameters of the virtual model and iteratively re-simulate the peeling process and product flow until at least one simulated output from the virtual model matches at least one desired output of the peeling system within a specified tolerance;

[0014] applying adjusted control parameters from the simulation of the virtual model to the peeling system in real time;

[0015] advancing the product to be peeled through a peeling system;

[0016] Monitor product and peeling processes and compare results with simulations;

[0017] Any identified differences between the measured output of the peeling system and the output of the virtual model are used to refine the virtual model to obtain a more accurate representation of the output of the peeling system.

[0018] This is advantageous because data about the product flow through the peeling system and / or about the product itself can be obtained in real time. Such data can be processed or simulated and provided to a virtual model of the peeling system, allowing the operation of the peeling system to be adjusted in real time based on the data. This provides for real-time optimization of the operation of the peeling system (i.e., the physical peeling system) based on data derived from the system in use. Those skilled in the art will recognize that the "output" of the peeling system as described can indicate the physical output of the peeling machine, as in the peeled product output. In addition, the "output" can also indicate any product data or system data output from sensors in the system. Real-time is given its general interpretation with respect to processes and process control, i.e., real-time refers to the actual time during which a process or event occurs, and specifically relates to a system that processes input data at a high frequency so that the input data is almost immediately available as feedback for the process from which it is derived.

[0019] In fact, an optimization iteration loop is provided in which simulations can be run and adjusted, and the output of the simulation is compared with the desired result (e.g., peeling quality or selected monitored system parameters) until the prediction from the simulation is considered close enough to the desired result. Once this is achieved, the resulting settings and parameters can be applied to the physical system. The virtual model of the system acts as a digital twin of the physical peeling system and provides a virtual or software environment in which the simulation of the physical system can be adjusted and analyzed, and further in which the performance and output of the physical system can be compared with the simulation. The output of the physical system can be monitored and compared with the simulation predictions from the "twin" in order to refine the simulation. The simulation can then be used to refine the performance of the physical system. In this way, visual inspections of the product line by people to determine product quality or peeling quality can be minimized. In addition, the downtime of the peeling system is also minimized because the system does not have to be stopped to perform inspections or system adjustments. This produces an energy- and cost-effective peeling system that improves peeling quality and is more reliable and has less system downtime.

[0020] The virtual model may include an optimizer and at least one of: a simulator for representing the operation of the peeling system; a simulated product flow for simulating the effect of the peeling system on the product; and a simulated product for simulating the effect of the product on the operation of the peeling system and / or the effect of the peeling system on the product.

[0021] This is advantageous because the simulator and optimizer can iteratively work together to optimize one or more outputs of the virtual model. This in turn provides for comparing the simulated outputs of the virtual model with the outputs of the peeling system (i.e., the physical system). Thus, an opportunity is provided to compare the simulated outputs with the "real world" outputs of the physical system. This allows the operation of the peeling system to be adjusted based on the difference between the ideal outputs of the simulator and the physical outputs of the peeling system.

[0022] The optimizer can be further configured to:

[0023] i) Analyse the output of the simulator and its effect on the simulated product and / or the effect of the peeling system on the product;

[0024] ii) comparing the simulated output of the simulator with a pre-specified output; and

[0025] iii) adjusting the control parameters to minimize the difference or error between the simulated output and the pre-specified output;

[0026] iv) iteratively repeating steps i) to iii) to provide optimized control parameters, wherein the optimized control parameters are determined when a difference or error between the simulated output and a pre-specified output is within a predetermined tolerance, the pre-specified output being defined by a desired product specification.

[0027] The method may further include applying the optimized control parameters to the peeling system.

[0028] This provides optimization of the operation of the peeling system as defined by the determined optimal control parameters. Optimization can be thought of mathematically as minimizing the difference or error between the desired output and the simulated output of the peeling system, typically through iterative minimization steps. This provides a type of model predictive control for the operation of the peeling system.

[0029] The control parameter may affect adjustment of at least one component of the peeling system and may include adjustment of at least one of the following of the peeling system:

[0030] Input product volume control;

[0031] Primary peeler;

[0032] secondary peeler;

[0033] loose rind separator;

[0034] Cleaning the system; and / or

[0035] Product sorting system. Each of the above elements is a component of a peeling system, however, those skilled in the art will appreciate that variations of these components are possible. Thus, it is advantageous to provide for adjustment of at least one of the components of the peeling system, as any such adjustment will have a significant effect on the peeling quality of the peelable product flowing through the system.

[0036] The control parameters affect at least one of the following:

[0037] Adjust product flow through the peeling system;

[0038] Adjusting the temperature at the primary peeler by applying heat or a heat-carrying medium;

[0039] adjusting the temperature at the secondary peeler by applying heat or a heat-carrying medium;

[0040] Adjust the peeling force at the first peeler;

[0041] Adjust the peeling force at the secondary peeler;

[0042] Adjust the sorting control of the outer skin;

[0043] Adjust defect sorting control;

[0044] Adjusting sorting controls to direct specific volumes of product;

[0045] Adjust the steam pressure in the system;

[0046] adjusting container temperatures at primary and / or secondary peelers;

[0047] Adjust vessel agitation at primary and / or secondary peelers;

[0048] adjusting the rotational speed of the primary and / or secondary peelers;

[0049] Adjust the number of rotation cycles at the primary and / or secondary peelers.

[0050] Each of the above parameters is a parameter to be defined when seeking a specific peeling quality at the output of the peeling system. Thus, it is advantageous to provide adjustment of at least one of these parameters as it will have a significant impact on the peeling quality of the peelable product flowing through the peeling system.

[0051] The characteristics of the product include the peelable skin coverage and quality of the product. Visual inspection of the product can provide an understanding of the skin quality of the product and thus an estimate of the difficulty of peeling. Skin quality can also be assessed by type and age or time in storage, possibly storage conditions. This information can be provided to the peeling system by the plant operator or through the plant process control network.

[0052] Additional characteristics of the product include coverage of product non-peelable defects, product size and / or surface area and / or volume. Information about what is peelable and what is non-peelable means that the peeling system will not be misadjusted for the peel it cannot remove, resulting in over-peeling and product loss. It should be noted that the term "non-peelable" can refer to peels that cannot be peeled by the primary peeler if it is directed to a secondary peeler that can remove it. In addition, it can also refer to non-peelable defects that need to be sent to waste because they cannot be peeled by the primary peeling device or the secondary peeling device. Additional characteristics of the product include product temperature. The peel is the surface condition, that is, more surface means more peel. Heat transfer is related to surface contact with the heat transfer medium (steam). Therefore, measuring heat transfer provides a direct relationship with the correct thermal energy to be applied. The temperature of the product is related to the initial amount of thermal energy required to cook its skin.

[0053] Additional characteristics of the product include the product category or type. Knowing the product category can provide information about other parameters of the peelability and performance of that category in a peeling system.

[0054] Each of the above is a significant characteristic of the peelable product itself, which in turn will directly affect the effectiveness of any peeling process applied to the peelable product. Thus, providing at least a first sensor for monitoring at least one characteristic of the product is advantageous because it provides a virtual model of the peeling system that can provide such data related to the product, directly related to the product quality, to optimize the operation of the peeling system.

[0055] The product flow may include at least one of the following:

[0056] Product flow rate;

[0057] product flow volume and / or mass;

[0058] product flow duration, pauses or dwell times;

[0059] Product flow piece counting.

[0060] These are significant characteristics of the product flow through the peeling system.Again, providing at least a first sensor for monitoring at least one characteristic of the product flow is advantageous as it provides that such data can be provided to a virtual model of the peeling system to optimize the operation of the peeling system.

[0061] The output of the virtual model provides an indication of the predicted peeling quality and / or defect removal quality of the product. This allows the operation of the peeling system to be adjusted based on the difference between the simulated peeling quality derived from the output of the virtual model and the observed physical output of the peeling system.

[0062] The output of the virtual model may be generated by applying machine learning and / or artificial intelligence to at least one of the following:

[0063] data from at least a first sensor for simulating a peeling process and product flow;

[0064] Historical data of peeling system;

[0065] Data from additional systems external to the peeling system;

[0066] Data from additional simulations outside the peeling system;

[0067] Data from cloud storage and / or cloud computer systems.

[0068] Within the scope of the present invention, artificial intelligence is understood as a technology that enables machines to simulate human behavior. Machine learning is a subset of artificial intelligence that allows machines to automatically learn from past data without explicit programming. Therefore, references to these terms should be understood in their broad sense and include learning techniques such as deep learning and / or neural networks. In this way, the virtual model can perform a prediction of the operation of the peeling system. This is advantageous because the prediction can be used to define the operation of the peeling system in a variety of ways. For example, initial startup data can be provided to the virtual model, and a predicted output of a product or parameter can be provided based on the provided data. The parameters of the peeling system can be adjusted so that the predicted output is "matched" with the physical output within a specified tolerance. In addition, the predicted output can provide an indication of an error or fault within the peeling system, where the predicted output is observed to deviate significantly from the output of the physical system. In this way, the predicted output can provide preventive fault management, in which the suboptimal physical operation of the peeling system components is revealed before the components are further degraded or damaged. Thus, the output of the virtual model can provide a fault indication or alarm. Likewise, this provides the advantage of minimizing system downtime for repairing or replacing parts. Furthermore, the use of historical data is advantageous because the virtual model thus provides information related to product flows that have previously passed through the peeling system. This data may, for example, include data on peelable products that have similar characteristics to the next proposed product to be loaded into the peeling system. Thus, the virtual model can simulate the product flow through the peeling system based on historical inputs. Likewise, the output of the simulation can be analyzed to verify whether it is within an acceptable quality range. If so, initial operating conditions can be configured for the peeling system based on historical data. This is advantageous because it provides that in scenarios where data derived from similar product flows through the peeling system are available, the operating conditions do not have to be determined from scratch. Using data from additional systems external to the peeling system, data from additional simulations external to the peeling system, or data from cloud storage similarly has advantages.

[0069] In another aspect of the present invention, there is provided a method of controlling a peeling system, comprising:

[0070] The virtual

[0071] The pseudo-model provides inputs including at least one initial operating condition;

[0072] The product flow through the peeling system is simulated in the virtual model based on the inputs to obtain

[0073] simulation;

[0074] Analyze the output of the simulation to obtain the analysis;

[0075] At least one operating condition is adjusted based on the analysis to produce an updated initial operating condition

[0076] condition;

[0077] Provide updated initial operating conditions to the peeling system.

[0078] This is advantageous because initial operating conditions for the peeling system can be provided to a virtual model of the peeling system to provide a simulated output. This simulated output can be analyzed to verify whether it is within an acceptable quality range. The operating conditions can be adjusted until the output has a sufficient desired quality. Thus, such adjusted initial operating conditions can be provided to the peeling system before the peeling process begins. In this way, the peeling system is set up or calibrated before use to achieve a target quality based on the simulated output.

[0079] In a further aspect of the invention, a peeling system is provided, comprising: a primary peeler for receiving a peelable product; a first sensor configured to monitor the peelable product at the primary peeler; wherein the first sensor is further configured to provide monitoring information to a virtual model of the peeling system for adjusting the operation of the peeling system based on the monitoring information. In this way, the operation of the peeling system can be adjusted based on information provided from the sensor at the primary peeler to the virtual model of the peeling system.

[0080] The peeling system may further include: a secondary peeler configured to receive a partially peeled product from the primary peeler; a second sensor configured to monitor the peelable product at the secondary peeling device; wherein the first sensor and the second sensor are configured to provide monitoring information to the virtual model of the peeling system for adjusting the operation of the peeling system based on the monitoring information. In this manner, the operation of the peeling system may be adjusted based on information provided to the virtual model of the peeling system from the first sensor at the primary peeler and the second sensor at the secondary peeler.

[0081] In another aspect of the present invention, there is provided a peeling system comprising:

[0082] Means for monitoring the peeling process of a peeling system and products peeled by the peeling system in real time using:

[0083] at least a first sensor for monitoring at least one characteristic of the product and for monitoring the operation of the peeling process;

[0084] Means for obtaining data from at least a first sensor to simulate in a virtual model a peeling process and a product flow through the peeling system to obtain a simulation, the simulation comprising:

[0085] means for comparing at least one simulated output of the virtual model with at least one desired output of the peeling system using an optimizer of the virtual model;

[0086] means for adjusting control parameters of the virtual model using the output of the optimizer and iteratively resimulating the peeling process and product flow until the output from the virtual model matches at least one desired output within a specified tolerance;

[0087] means for applying adjusted control parameters from the simulation of the virtual model to the peeling system in real time;

[0088] means for advancing the product to be peeled through the peeling system;

[0089] a device for monitoring the product and peeling process and comparing the results with simulations;

[0090] Means for utilizing any identified differences between the measured output of the peeling system and the output of the virtual model to refine the virtual model to more accurately represent the output of the peeling system.

[0091] Such a control system provides data available in real time about the product flow through the peeling system and / or about the product itself. Such data can be processed or simulated and provided to a virtual model of the peeling system, thereby allowing the operation of the peeling system to be adjusted based on the data. BRIEF DESCRIPTION OF THE DRAWINGS

[0092] Figure 1 is a representation of the peeling process flow performed by the peeling system.

[0093] Figure 2 is a schematic representation of the model predictive control (MPC) process.

[0094] Figure 3 is a schematic representation of the model predictive control (MPC) process used to control the peeling system.

[0095] Figure 4a is a schematic representation of a peeling system.

[0096] Figure 4b is a schematic representation of a peeling system according to the present invention further indicating monitoring and control aspects. DETAILED DESCRIPTION

[0097] Before describing the functionality of the present invention, a brief description of a typical peeling process and system is provided. Figure 1 is a representation of the peeling process flow performed by the peeling system.

[0098] The steps performed during the peeling / sorting stage are summarized in more detail below.

[0099] A-Startup Batch

[0100] Step A includes first assessing the size, temperature, skin type, and defects of the raw material (e.g., potatoes). The batch size is selected based on an estimate of the size distribution and throughput constraints of the system (i.e., how fast the product can flow through a given system). The steam pressure and temperature available in the steam vessel are assessed. The peeling time is set based on the assessment of the raw material, batch size, and steam temperature information. The batch is loaded into the steam vessel and the cycle is started.

[0101] B-Peeling batch

[0102] The steam vessel acts as a primary peeler for the loaded batch. The vessel rotates the potatoes in a pressurized steam environment. The pressure is assessed throughout the cycle. At the end of the set peeling time, the dump cycle is performed and the vessel is rotated to dump the potatoes.

[0103] C-Clean Batch

[0104] The potatoes are conveyed to the DPS via an auger discharger. The potatoes travel through the DPS to remove the loose skin of the potatoes. Brushes remove the remaining attached skin. The product is then distributed on a hex-feed to remove any product debris and spread the product evenly for sorting.

[0105] D-sorting

[0106] The sorter removes foreign matter and defective products into waste bins and sorts the remaining products into “in specification” and “out of specification”.

[0107] E - Re-peeling batches

[0108] The secondary peeler peels all the products received from the sorter. The peeled output of the secondary peeler is re-sorted through the channel in the sorter. The under-peeled output of the secondary peeler is re-peeled again after sorting.

[0109] F-Finish

[0110] The output product is "on specification" and a final inspection and grading of the output product is performed to ensure it is of the required quality.

[0111] There are multiple problems with the control of such systems. In existing peeling systems, the sorter and steam container may be controlled by a powertrain control module (PCM) system. Such systems provide limited control and visibility of system parameters and product characteristics that determine peeling quality. For example, it is not until the quality control stage after peeling / sorting that it is possible to know whether the product is determined to be within or outside the specification. The peeling / sorting stage is also unable to cope with the full range of input product quality without significant adjustments. Sorter control can also be independent of peeling control. Peeling efficiency also depends on the quality of steam at the steam container, but throughout the process, the quality of the steam may still not be monitored and ultimately unknown. In addition, the incoming product specifications, product batch size, and steam container temperature have a mutually dependent relationship, which affects the peeling quality. The current control system has not yet assessed this dependency. The method of the present invention attempts to address such deficiencies.

[0112] Figure 2 is a schematic representation of a system (e.g., a peeling system) implemented using a control method according to the present invention. The method utilizes a form of model predictive control (MPC). MPC is a method for controlling a system or process while satisfying a set of constraints. A model predictive controller relies on a dynamic model or virtual model of the controlled process. Data obtained from operating the process is provided to the model, which is used to generate a simulated output. The simulated output can be compared with the actual output to adjust the control parameters of the system.

[0113] about Figure 2 , where the process 20 represented can be considered as being about Figure 1 and as further described below Figure 4A and 4B Describe the peeling / sorting process.

[0114] The method of the present invention provides real-time monitoring of the peeling and / or sorting process of the system and of the products peeled and sorted by the system using at least a first sensor 21 for monitoring at least one characteristic of the product and for monitoring the operation of the process;

[0115] Multiple sensors can be placed at different points throughout the system (see, for example, Figure 4B) to collect relevant process status data about the operation of the peeling / sorting equipment and the products flowing through the equipment. The detected data is provided to a virtual model 22 of the peeling system. The virtual model 22 is shown as including an optimizer 23 and a simulator 24 or simulation model. The simulator 24 is used to represent the operation of the peeling and sorting system. In addition, simulated products are provided (but not shown in this figure) to simulate the impact of the product on the operation of the peeling system and the impact of the system on the product. In addition, a simulated product flow (not shown) is provided to simulate the impact of the peeling system on the product. The product flow describes the movement of the product through the peeling system. The product flow is defined by multiple factors, including but not limited to product flow rate; product flow volume and / or mass; product flow duration, pause or residence time; piece count of the product flow.

[0116] The data obtained from at least the first sensor is used to simulate the peeling process and the product flow through the peeling system in a virtual model. The simulation includes comparing at least one simulated output of the virtual model 22 with at least one desired output of the peeling system using an optimizer 23. The control parameters of the virtual model are adjusted using the output of the optimizer. The peeling process and product flow are then iteratively re-simulated until the output from the virtual model matches the at least one desired output within a specified tolerance.

[0117] The adjusted control parameters from the simulation of the virtual model are applied to the peeling system in real time. The product flow of the product to be peeled then travels through the peeling system. The product and peeling process are monitored and the results are compared to the simulation. Any identified differences between the measured output of the peeling system and the output of the virtual model 22 are used to refine the virtual model to more accurately represent the output of the peeling system. The measured output of the peeling system can be the physical output of the peeling machine, as in peeling the product. In addition, the measured output can be any product or data output from the system or sensors in the system.

[0118] The optimizer 23 may also be provided with external inputs from an operator or external controls 25. External inputs may include, for example, output specifications, specific heat values, and dry solids content. Figure 3 It is shown how an information loop is created to provide process data from the physical equipment (i.e., data detected from sensors distributed throughout the system) to the simulator 24. A further loop is created by providing updated simulation predictions from the simulator 24 to the optimizer 23. This in turn allows the optimizer 23 to provide updated control actions to the simulator 24, which ultimately translate into updated control parameters for the physical equipment.

[0119] The optimizer is configured to: i) analyze the output of the simulator and its impact on the simulated product; and ii) compare the simulated output of the simulator with the pre-specified output. The optimizer can then iii) adjust the control parameters to minimize the difference or error between the simulated output and the pre-specified output.

[0120] Furthermore, the optimizer is configured to iteratively repeat steps i) to iii) to provide optimized control parameters. The optimized control parameters are determined when the difference or error between the simulated output and the pre-specified output is within a predetermined tolerance. The pre-specified output is defined by a desired product specification. For example, the pre-specified output may be a peeling quality for a given peelable product. The pre-specified output may also be a data output of any aspect of the peeling system. The optimized control parameters may then be applied to the peeling system.

[0121] In addition, adjusting the control parameters 24 may include minimizing the error between the output of the simulator and the target output of the peeling system. The output may be a selected parameter or threshold value defined for any process part as listed in A to F above. The output of the virtual model is generated by applying machine learning and / or artificial intelligence to at least one data from a sensor in the peeling system to simulate the peeling process and product flow. The output of the virtual model may also be generated by applying machine learning and / or artificial intelligence to historical data from the peeling system, data from an additional system outside the peeling system (e.g., another peeling system in a different production environment), data from additional simulations outside the peeling system, and data from cloud storage. In addition, the output of the virtual model is configured to provide an overall indication of the peeling quality of a given peelable product. 22 The virtual peeling quality can be compared with the output of the physical system. For example, the physical output at the sorter 13 can be inspected for overall peeling quality, and whether the output product is within or outside the specification. In addition, the physical output can be compared with the predicted simulation output (i.e., virtual peeling quality) from the virtual model 22. In this way, the accuracy of the predicted simulation output can be determined. The simulator 24 may be adjusted, for example, by manual input, to adjust the simulated output based on the observed physical output.

[0122] Note that providing the information loop described provides for iterative adjustments to the operation of the peeling system.

[0123] Figure 3 is a schematic representation of a control method for controlling a peeling system. Figure 3 Instructions on how to Figure 3 The control method described is applied to Figure 1 and Figure 2The peeling / sorting process described in . In practice, the various controls and control parameters described can be applied to multiple steps A to F of the peeling / sorting process, i.e., batch size control is applied to step A - starting the batch; steam time control is applied to step B - peeling the batch; sorter and ejection control is applied to step D - sorting the batch, and re-peeling control is applied to step E - re-peeling the batch. This actually provides a series of distributed MPC controls placed throughout the process, with a master MPC providing overall control and supervision. User input can also be applied to the master control as required. In addition, the overall production line control or plant supervision can be linked to the master control. In this way, centralized control and monitoring of the entire production line can be provided.

[0124] Applying batch size control at step A provides enhanced monitoring and iterative control of multiple process aspects performed at step A. For example, additional monitoring and evaluation can be provided by sensors configured to evaluate raw material size temperature, crust type and defects. Additional monitoring and evaluation can be provided by sensors configured to evaluate, for example, available steam pressure and temperature and measure container temperature. In addition, applying batch size control can allow dynamic adjustment of the process based on data from the process, model simulations, and desired product specifications. Adjustments can be made to the primary peeler. In addition, adjustments to batch size and volume can be provided. Adjustments to the peeling time of the batch can also be provided.

[0125] Applying steam time control at step B provides enhanced monitoring and iterative control of multiple process aspects performed at step B. For example, additional monitoring and evaluation may be provided by sensors configured to open a steam inlet to a steam vessel, pressurize the vessel, and measure the steam pressurization cycle. Further data may be obtained from sensors configured to release steam pressure and measure a steam release profile as the steam is discharged.

[0126] The application of the sorter and discharge control at step D provides enhanced monitoring and iterative control of multiple process aspects performed at step D. For example, additional monitoring and evaluation can be provided by sensors configured to obtain data regarding defects identified from images taken from the output of steps A to D, obtain data regarding the number of products sent to the secondary peeler, and send data regarding the size and defects of products sent to the re-peeling stage. In addition, the application of the sorter and discharge control can allow dynamic adjustments to the process to be provided. For example, adjustments to the sorter operation can be provided in relation to determining which products are sorted as in-spec, out-of-spec, and discarded. Adjustments to the brush operation can also be provided.

[0127] Applying repeeling control at step E provides enhanced monitoring and iterative control of multiple process aspects performed at step E. For example, additional monitoring and evaluation can be provided by sensors configured to obtain data regarding the effects of repeeling. Furthermore, applying repeeling control can allow for dynamic adjustments to the process based on data from the process, model simulations, and desired product specifications. For example, adjustments to the operation of a secondary peeler can be provided.

[0128] Note that in addition to the adjustments described, additional adjustments may be made to the system in order to optimize the operation of the peeling system. Ultimately, multiple aspects of the peeling system and process may be controlled and optimized by adjusting the control parameters. Adjusting the control parameters may affect adjustment of at least one of the following items of the peeling system: input product volume control; primary peeler; secondary peeler; loose rind separator; cleaning system; product sorting system.

[0129] Furthermore, it should be noted that sensors can be introduced at various locations throughout the peeling system in order to obtain data for process optimization. The sensors can be configured to obtain data related to the product itself, for example, obtaining data indicating the percentage of product defects, product temperature, product surface area, product skin thickness. Furthermore, the sensors can be configured to obtain data related to the product flow, for example, obtaining data indicating the product flow velocity in the system, the product flow duration, the product volume, or the system throughput. Furthermore, the control parameters affect adjustments of: adjustment of the product flow through the peeling system, adjustment of the temperature at the primary peeler by applying heat or a heat carrier medium, adjustment of the temperature at the secondary peeler by applying heat or a heat carrier medium, adjustment of the peeling force at the first peeler, adjustment of the peeling force at the secondary peeler, adjustment of the sorting control of the rind, adjustment of the sorting control of defects, adjustment of the sorting control for directing a specific volume of product, where directing may mean redirecting the product through the system or discharging it from the product flow, steam pressure in the system, container temperature at the primary peeler and / or secondary peeler, container agitation at the primary peeler and / or secondary peeler, rotation speed of the primary peeler and / or secondary peeler, number of rotation cycles at the primary peeler and / or secondary peeler.

[0130] It should be noted that aspects of the described method can be further applied before starting the peeling process so that the initial conditions of the peeling process for a given peelable product are optimized. Thus, an operator or external control can provide inputs including at least one initial operating condition to a virtual model of the peeling system. The initial operating conditions can be operating parameters of components of the peeling system, such as operating parameters of a primary peeler, a secondary peeler, or a sorter. The optimizer and simulator combination can then simulate the product flow through the peeling system based on the inputs; and then analyze the output of the simulation. The output can be, for example, an indication of the overall peeling quality. If the peeling quality is not within the specification, then the initial operating conditions can be adjusted and updated based on the analysis to provide an optimized simulated output within the specification. Then, this updated initial operating condition can be provided as an input to the physical peeling system so that the system is set or calibrated based on the simulation to provide an output of the desired specification. Once the process is ongoing in the peeling system, if, for example, feedback indicates that the output deviates from the specification, then real-time adjustments can be made as previously described.

[0131] The input to the peeling system may also be in the form of historical data of product flow through the peeling system. This historical data may be used to actually prepare the peeling system to operate in a manner based on the product flow that has occurred. For example, if the product parameters of the historical product flow appear similar to the parameters of a new product to be loaded into the system, then the system parameters for the new product may be initialized based on the parameters for the historical flow for the similar product.

[0132] Figure 4A is a schematic representation of a peeling system. This representation provides an overview of one type of peeling system suitable for control by the method of the present invention. Figure 4B Before describing the various aspects of the present invention that provide enhanced monitoring and control, such a peeling system will be briefly described.

[0133] During the peeling process, the products go through several preparation stages that transform these products from raw products to final packaged products suitable for sale and consumption. Broadly speaking, these stages include: Receiving raw products -> cleaning / sizing / and blending -> Peeling / Sorting -> cutting -> strip sorting -> blanching -> battering -> frying -> freezing -> sorting -> packaging. Figure 4A The representation shows the use of peelable products Peeling / SortingEquipment for the stage. The operation of the peeling and sorting process will be briefly described. Multiple problems of the existing control methods of such systems will also be described. Note that, although the peeling and sorting process performed by the peeling system will be described with respect to potatoes, the process and system can be used for peeling and sorting other types of peelable products and peelable foods. Thus, it should not be considered that the described method and system are only for potatoes. Different types of peelable products can be peeled by the peeling system. The product has multiple characteristics, including but not limited to the coverage of the peelable skin of the product, the quality of the peelable skin of the product, the coverage of the non-peelable defects of the product, the size and / or surface area and / or volume of the product, the temperature of the product, the type or type of the product. These characteristics are used to define the type of product that is undergoing peeling through the system. In addition, a peelable product refers to any product that can be peeled to remove the outer layer of the product, especially food. The peelable product may include an unpeeled product, a partially peeled product, or a completely peeled product. In addition, "product" refers to one or more of a given item, but more generally a plurality of product items, such as thousands of potatoes or other peelable foods.

[0134] Feed control

[0135] This part of the process utilizes a metering conveyor or product transport mechanism and a load cell. Washed, unpeeled potatoes are fed from the conveyor into a batch hopper. A volumetric measurement system may also be employed to measure the fill depth to estimate the volume of product to be processed.

[0136] Primary peeler - active peel separation

[0137] From the batch hopper, the potatoes arrive at the steam vessel, which operates as the primary peeler 11 of the system. High pressure steam is fed into the steam vessel, which, combined with rapid decompression, provides active peel separation. Once the peeling cycle is complete, a steam exhaust removes the steam from the vessel.

[0138] Remove loose and attached skins and starch

[0139] The unloader screw conveys the potatoes from the steam receiver to the Dry Skin Separator (DPS). The peel separator and brush / scrubber equipment removes loose and partially attached peels from the product prior to inspection.

[0140] Sorting and secondary peeling machines

[0141] The potatoes are evenly distributed and free of debris and fed into the sorter 13 device via hexagonal rollers. The device is configured to sort the potatoes into i) peeled potatoes within the required specifications or peeling quality ("in specification"), and ii) peeled potatoes that do not meet the required specifications or peeling quality or contain defects that can be removed by a secondary peeler ("out of specification"), and iii) defective potatoes and any detected foreign matter. Defective potatoes and foreign material are discarded, peeled potatoes within specifications are directed out of the sorter 13, and peeled potatoes that do not meet the required specifications are directed to the secondary peeler 12. At the secondary peeler 12, the remaining outer skin and surface defects can be removed. The re-peeled potatoes are then redirected back to the sorter 13, where they can be sorted again according to i), ii) or iii) as described above. In some applications, this re-peeled product can be directly combined with the product that meets the specifications without re-inspection and sorting.

[0142] Figure 4B Several enhancements are shown that provide advantages over Figure 4A Control and monitoring of the system shown in . Figure 4B The method for controlling a peeling system of the present invention uses at least a first sensor to monitor the peeling process of the peeling system and the products peeled by the peeling system in real time, and the at least first sensor is used to monitor at least one characteristic of the product and monitor the operation of the process.

[0143] refer to Figure 4B, it is noted that at least one sensor may be placed at one of a number of different points throughout the peeling system, thus providing the ability to monitor product characteristics and / or the operation of the different equipment components that make up the peeling system. For example, sensors may be deployed to monitor various aspects of washed unpeeled potatoes. For example, sensors may be deployed and configured to monitor input specifications of incoming products. These may be used to define a "raw product score" or initial product classification. In addition, the input specifications of the incoming product may be used to define feedforward control parameters. One or more sensors may be placed around the metering conveyor. This may provide information about the weight / mass of the product, the level or volume of the product, and the temperature of the product. One or more sensors may be placed around the steam vessel. This may provide information about steam pressure, valve operation, and vessel temperature. Again, such information may be used to determine and / or adjust control parameters. One or more sensors may also monitor the weight / mass of the product and the volume of the product at the screw discharger. One or more sensors may also be deployed at a checkpoint after the peeling separator. One or more sensors may be placed around the sorter and secondary peeler. These sensors provide data about the product from the sorter, including peelable and non-peelable defects. Furthermore, information on the performance of the primary and secondary peelers is available. This allows the definition and adjustment of parameters for controlling the classifier program, such as the classifier settings. Sensors at the secondary peeler provide a smarter monitoring of the secondary peeler performance. This provides management of the re-peeling load and control of the product direction.

[0144] It should be noted that the locations of the sensors described are only one indication of possible sensor locations. One or more sensors may be placed throughout the system to obtain different types of information depending on the type of system performance to be monitored. Furthermore, it should be noted that a single sensor may be able to monitor more than a single system aspect and may be able to obtain more than one data type.

[0145] It has been noted that the data obtained from the sensors provides a simulation to simulate the peeling process and the product flow through the peeling system in a virtual model. The obtained sensor data can be used to Figure 2 The model predictive control process described utilizes to refine the virtual model to more accurately represent the output of the peeling system. Figure 2 The highlighted process 20 can be performed by Figure 4B The system outlined in .

[0146] In addition to the described control method, a further embodiment provides a peeling system comprising: a primary peeler for receiving a peelable product; a first sensor configured to monitor the peelable product at the primary peeler; wherein the first sensor is also configured to provide monitoring information to the described virtual model for adjusting the operation of the peeling system based on the monitoring information.

[0147] The secondary peeler may also be configured to receive a partially peeled product from the primary peeler. As described, the partially peeled product may also be provided to the secondary peeler from the sorter. The second sensor is configured to monitor the peelable product at the secondary peeling device. The first and second sensors are configured to provide monitoring information to a virtual model of the peeling system for adjusting the operation of the peeling system based on the monitoring information. In addition, the third sensor may be configured to monitor the peelable product at the sorter, and the first, second, and third sensors are configured to provide monitoring information to the virtual model of the peeling system for adjusting the operation of the peeling system based on the monitoring information.

[0148] The method as described can be controlled by a computer system and furthermore the method can be implemented in software or any computer program product comprising instructions which, when executed by a computer, cause the computer to perform the control method as described. The method as described can be performed by a control system integrated with a physical peeling system comprising mechanical, electrical, electronic and software means and combinations thereof configured to perform the method. The control system for the peeling system comprises means for real-time monitoring of the peeling process of the peeling system and the product peeled by the peeling system using at least a first sensor for monitoring at least one characteristic of the product and monitoring the operation of the process.

[0149] The control system further comprises means for obtaining data from at least the first sensor to simulate in a virtual model the peeling process and the product flow through the peeling system, the simulation comprising:

[0150] means for comparing at least one simulated output of the virtual model with at least one desired output of the peeling system using an optimizer of the virtual model;

[0151] Means for using the output of the optimizer to adjust control parameters of the virtual model and iteratively resimulate the peeling process and product flow until the output from the virtual model matches at least one desired output within specified tolerances.

[0152] The control system also includes means for applying adjusted control parameters from the simulation of the virtual model to the peeling system in real time; means for advancing the product to be peeled through the peeling system;

[0153] Means for monitoring the product and peeling process and comparing the results with the simulation; means for utilizing any identified differences between the measured output of the peeling system and the output of the virtual model to refine the virtual model to more accurately represent the output of the peeling system.

[0154] It has been described how existing control methods for peeling systems provide limited control over the system and the peeling process performed by the system. Furthermore, it is difficult to see whether the process is adequately performing within operating parameters to produce an output product that meets desired specifications. The present method as described provides increased system uptime and increased peeling efficiency. This in turn provides maximized product yield and reduced waste.

[0155] The words “comprises / comprising” and the words “having / including” used herein with respect to the present invention are used to specify the presence of stated features, integers, steps or components, but do not exclude the presence or addition of one or more other features, integers, steps, components or groups thereof.

[0156] It should be appreciated that, for the sake of clarity, certain features of the present invention described in the context of separate embodiments may also be provided in combination in a single embodiment. Conversely, for the sake of brevity, various features of the present invention described in the context of a single embodiment may also be provided separately or in any suitable sub-combination.

Claims

1. A method for controlling a peeling system, comprising: Monitor the peeling process of the peeling system and the products peeled by the peeling system in real time using: at least a first sensor for monitoring at least one characteristic of the product, and Used to monitor process operations; Obtaining data from at least a first sensor to perform a simulation, the simulation simulating a peeling process and a product flow through a peeling system in a virtual model, the simulation comprising; comparing at least one simulated output of the virtual model to at least one desired output of the peeling system using an optimizer of the virtual model; using the output of the optimizer to adjust control parameters of the virtual model and iteratively re-simulate the peeling process and product flow until the at least one simulated output from the virtual model matches the at least one desired output of the peeling system within a specified tolerance; applying adjusted control parameters from the simulation of the virtual model to the peeling system in real time; advancing the product to be peeled through a peeling system; Monitor product and peeling processes and compare results with simulations; Any identified differences between the measured output of the peeling system and the output of the virtual model are used to refine the virtual model to obtain a more accurate representation of the output of the peeling system.

2. The method of claim 1, wherein the virtual model comprises an optimizer and at least one of: a simulator to represent the operation of the peeling system; Simulated product flow to simulate the impact of the peeling system on the product; Simulated product used to simulate the effect of the product on the operation of the peeling system and / or the effect of the peeling system on the product.

3. The method of claim 2, wherein the optimizer is further configured to: i) Analyze the output of the simulator and its impact on the simulated products and / or product flows; ii) comparing the simulated output of the simulator with a pre-specified output; and iii) adjusting the control parameters to minimize the difference or error between the simulated output and the pre-specified output; iv) iteratively repeating steps i) to iii) to provide optimized control parameters, wherein, Optimized control parameters are determined when the difference or error between the simulated output and a pre-specified output, which is defined by the desired product specifications, is within a predetermined tolerance.

4. The method of claim 3, further comprising: Apply the optimized control parameters to the peeling system.

5. A method as claimed in any preceding claim, wherein the control parameter affects adjustment of at least one of the following of the peeling system: Input product volume control; Primary peeler; secondary peeler; loose rind separator; Cleaning the system; and / or Product sorting system.

6. The method of claim 5, wherein the control parameter affects at least one of: Adjust product flow through the peeling system; Adjusting the temperature at the primary peeler by applying heat or a heat-carrying medium; adjusting the temperature at the secondary peeler by applying heat or a heat-carrying medium; Adjust the peeling force at the first peeler; Adjust the peeling force at the secondary peeler; Adjust the sorting control of the outer skin; Adjust defect sorting control; Adjusting sorting controls to direct specific volumes of product; Adjust the steam pressure in the system; adjusting container temperatures at primary and / or secondary peelers; Adjust vessel agitation at primary and / or secondary peelers; adjusting the rotational speed of the primary and / or secondary peelers; Adjust the number of rotation cycles at the primary and / or secondary peelers.

7. A method as claimed in any preceding claim, wherein the product stream comprises at least one of: Product flow speed; product flow volume and / or mass; product flow duration, pauses or dwell times; Product flow piece counting.

8. A method as claimed in any preceding claim, wherein the characteristics of the product include at least one of the following: The peelable skin coverage of the product; The quality of the peelable skin of the product; Defect coverage for products that cannot be peeled; product size and / or surface area and / or volume; Product temperature; Product category or type.

9. A method as claimed in any preceding claim, wherein the output of the virtual model provides an indication of the predicted skin removal quality and / or defect removal quality of the product.

10. A method as claimed in any preceding claim, wherein the output of the virtual model is produced by applying machine learning and / or artificial intelligence to at least one of: data from at least a first sensor for simulating a peeling process and product flow; Historical data of peeling system; data from at least one additional system external to the peeling system; Data from additional simulations outside the peeling system; Data from cloud storage and / or cloud computing systems.

11. A control system for a peeling system, comprising: Means for monitoring the peeling process of a peeling system and products peeled by the peeling system in real time using: at least a first sensor for monitoring at least one characteristic of the product and for monitoring the operation of the process; means for obtaining data from at least a first sensor to perform a simulation, said simulation simulating in a virtual model a peeling process and a product flow through the peeling system, said simulation comprising; means for comparing at least one simulated output of the virtual model with at least one desired output of the peeling system using an optimizer of the virtual model; means for adjusting control parameters of the virtual model using the output of the optimizer and iteratively resimulating the peeling process and product flow until the output from the virtual model matches at least one desired output within a specified tolerance; means for applying adjusted control parameters from the simulation of the virtual model to the peeling system in real time; means for advancing the product to be peeled through the peeling system; a device for monitoring the product and peeling process and comparing the results with simulations; Means for utilizing any identified differences between the measured output of the peeling system and the output of the virtual model to refine the virtual model to obtain a more accurate representation of the output of the peeling system.

12. The system of claim 11, further configured to perform the method of any one of claims 2 to 10.

Images