Method for controlling the operation of a sludge dehydration system, and associated control system

Abstract

The present invention relates to a method for controlling the operation of a sludge dehydration system, the method comprising the following steps: - acquiring (110) a plurality of characterisation parameters characterising output products; - determining (120) a control setpoint by applying fuzzy rules to the characterisation parameters according to fuzzy logic, the fuzzy rules being derived from a predetermined fuzzy rule base, each fuzzy rule being based on at least one input parameter independent of the dehydration system; - applying (130) the control setpoint to the dehydration system; wherein the determination step (120) comprises a sub-step (121) of preprocessing the characterisation parameters, the sub-step comprising the conversion of the characterisation parameters into input parameters.

Description

[0001] DESCRIPTION

[0002] TITLE: Method for monitoring the operation of a sludge dewatering system and associated control system

[0003] The present invention relates to a method for controlling the operation of a dehydration system.

[0004] The present invention also relates to a control system associated with such a process.

[0005] More specifically, the process and control system according to the invention make it possible to control the operation of a dewatering system treating sewage sludge.

[0006] Sewage sludge, as is known in itself, is a by-product of wastewater treatment by a treatment plant associated, for example, with a geographical region.

[0007] Over the past few decades, the quantities of sludge produced have increased considerably due to rapid growth in industrialization and population, as well as stricter wastewater treatment standards.

[0008] The annual production of sludge in Europe increased from 6.5 million tonnes of dry matter (DM) in 1992 to 13 million tonnes DM in 2020. Furthermore, a 10% increase in the total disposal of sludge in France is projected by 2050. This will lead to higher costs for the treatment, recovery and transport of sludge.

[0009] Therefore, improving the operational performance of sludge management requires the development of sustainable strategies aimed at reducing the amount of sludge to be treated, increasing its dryness (reducing its water content) to maximize material / energy recovery and minimizing its environmental impact.

[0010] In existing sludge management systems, mechanical dewatering processes are commonly used to reduce sludge volume, with the aim of optimizing its energy or agricultural recovery, as well as its storage and transport.

[0011] Screw centrifuge dewatering is one of the most widespread technologies in the sludge treatment process. It is generally used after chemical conditioning of the sludge to improve its dryness. A centrifuge aims to separate the sludge into a solid cake and a clarified liquid containing little suspended solids (SS). The performance of the dewatering process can be evaluated based on several factors, such as the final dry matter content of the dewatered sludge, the quality of the centrate, and the quantity of chemicals (polymers) added to the sludge to be dewatered.

[0012] Conventional methods of regulating dehydration by centrifugation often have limitations in terms of the operating stability of the centrifuges, the dryness of the filter cake, the rate of capture of solids, the optimal dosage of the polymer, etc.

[0013] Traditional dehydration regulations rely on measurements of suspended solids (SS) concentration and incorporate different strategies:

[0014] 1. Mass flow regulation: Measurement of the TSS concentration and hydraulic flow rate to adjust the feed pump flow rate. However, sensor reliability can be limited by sludge variations, disturbances such as fibrous material, or excessive discoloration.

[0015] 2. Proportional dosing of the reagent: Adjustment based on the incoming flow, with dosing based on assumptions of constant effluent concentration. This method can lead to overdosing.

[0016] 3. Turbidity measurement in the centrate: Adjustment of the reagent flow rate based on the measured turbidity, although this measurement may be disturbed by phenomena such as foams or microbubbles.

[0017] Variations in incoming sludge (concentration, quality) and polymer (concentration and efficiency) complicate regulation and often require manual intervention to adjust parameters.

[0018] Some of the monitoring and control methods proposed more recently rely on machine learning models based on data acquired in particular by sensors of different kinds.

[0019] However, the maintenance requirements of the sensors and the challenges associated with data-driven modeling (requiring both quantity and quality of data) limit the operational use of these solutions. In particular, for modeling to be effective and accurate, the system must include numerous sensors, which is very expensive. Furthermore, maintaining these sensors also presents cost and, sometimes, feasibility issues.

[0020] In this context, Nader Moatamri's document, "From the Analysis of Centrifuge Control to its Instrumentation," proposes a method for controlling a centrifuge based on sensor measurements. However, this method is adapted to a specific type of centrifuge, which limits its application to other types of centrifuges.

[0021] Document W01999028040A1 proposes a control solution for centrifuges used for solid / liquid separation, particularly for dewatering sewage sludge. This process optimizes performance while eliminating the need for human monitoring, thanks to the application of fuzzy logic. This control method is based on the continuous measurement of key variables (suspended solids, sludge flow rate, polymer flow rate, torque, or relative velocity). However, the failure of the suspended solids measurement system on which this innovation relies limits the application of this control method.

[0022] The present invention aims to remedy all the aforementioned problems and thus to propose means to effectively control a dehydration system even when it is completely devoid of sensors or when it is equipped with few sensors and / or low-quality sensors, regardless of the type of dehydration system.

[0023] To this end, the invention relates to a method for controlling the operation of a sludge dewatering system, the dewatering system being configured to treat sewage sludge by producing output products, the method comprising the following steps:

[0024] - acquisition of a plurality of characterization parameters characterizing output products;

[0025] - determination of a control instruction by applying fuzzy rules to the characterization parameters according to a fuzzy logic, the fuzzy rules being derived from a predetermined base of fuzzy rules, each fuzzy rule being based on at least one input parameter independent of the dehydration system;

[0026] - application of the control instruction to the dehydration system;

[0027] in which the determination step includes a sub-step of preprocessing the characterization parameters including the conversion of the characterization parameters into input parameters.

[0028] Applying fuzzy rules to determine control instructions significantly reduces the requirements for the quality of the input data used to define these instructions. Thus, even with limited data, poor-quality data, or subjective data (determined, for example, by an operator), it is possible to effectively control the operation of the dehydration system. Furthermore, fuzzy rules can be constructed to be independent of any particular type of dehydration system. Therefore, the invention can be applied to control the operation of any type of dehydration system. In general, fuzzy rules enable the implementation of so-called "fuzzy logic," which extends classical logic to approximate reasoning. Due to its numerical aspects, fuzzy logic differs from modal logics.Unlike Boolean algebra, fuzzy logic allows the truth value of a condition to range across a domain other than the {true, false} pair. In fuzzy logic, there are degrees of truth in satisfying a condition. Fuzzy logic assigns degrees of truth to a relation of the form 'x is closer to y than to z,' constructed and / or refined through learning. In general, fuzzy relations allow for the encoding of gradual, empirical, or typical knowledge, acquired directly or through heuristics, induction algorithms, and so on. The membership functions used can be multilinear (e.g., triangular, trapezoidal), sigmoid, Gaussian, etc.

[0029] According to other advantageous aspects of the invention, the method comprises one or more of the following features, taken individually or in all technically possible combinations:

[0030] - the output products include dewatered sludge and a centrate, and in which the characterization parameters include an estimation of the dryness of the dewatered sludge, an estimation of the color of the centrate and / or an estimation of the flocculation of the centrate;

[0031] - Estimates of the color and / or flocculation and / or dryness of the output products are determined by comparing the color and / or flocculation and / or dryness of the output products with that of the predetermined samples;

[0032] - the characterization parameters are determined by one or more low precision sensors and / or by an operator;

[0033] - fuzzy rules link at least certain values ​​of characterization parameters to actions consisting of modifying operating parameters of the dehydration system;

[0034] - the control instruction includes one or more actions determined by the fuzzy rules;

[0035] - the operating parameters include a sewage sludge flow rate and a polymer flow rate for treating the sewage sludge;

[0036] - the dehydration system is a centrifuge comprising a bowl and a conveying screw, and in which the operating parameters further include a relative rotational speed of screw / bowl and a screw / bowl torque;

[0037] - Each fuzzy rule is determined from the knowledge of the dehydration system experts; - The conversion of characterization parameters into input parameters is done using hyperparameters, each hyperparameter being associated with one of the characterization parameters and defined according to the dehydration system; - Each hyperparameter presents an extremum of the associated characterization parameter which is applicable to the dehydration system;

[0038] - the sub-step of preprocessing the characterization parameters further includes the determination of a parameter derived from at least some of the characterization parameters;

[0039] - said derived parameter is a substrate capture rate determined from estimates of centrate color and / or centrate flocculation;

[0040] - the determination step further includes a sub-step of applying fuzzy rules according to fuzzy logic using a Generalized Modus Ponens technique.

[0041] The invention also relates to a computer program comprising software instructions which, when executed by a computer, implement a process as defined above.

[0042] The invention finally relates to a system for controlling the operation of a sludge dewatering system, comprising technical means configured to implement the process as defined above.

[0043] The invention will become clearer upon reading the following description, given solely by way of non-limiting example, and made with reference to the drawings in which:

[0044] [Fig. 1] Figure 1 is a schematic view of a control system according to the invention, the control system allowing control of the operation of a sludge dewatering system;

[0045] [Fig. 2] Figure 2 is a schematic view of an example implementation of the dehydration system of Figure 1;

[0046] [Fig. 3] Figure 3 is a flowchart of a control process, the control process being implemented by the control system of Figure 1;

[0047] [Fig. 4] [Fig. 5] [Fig. 6] Figures 4 to 6 present an example of a fuzzy rule and its application during the implementation of the process in Figure 3.

[0048] Figure 1 illustrates a control system 10 allowing control of the operation of a dehydration system 12, such as illustrated for example in Figure 2.

[0049] More specifically, the dewatering system 12 is, for example, part of a wastewater treatment plant. The plant is, for example, associated with a specific geographical area, such as a municipality or a group of municipalities.

[0050] The dewatering system 12 is configured for example to treat sewage sludge after chemical conditioning of this sludge within the treatment plant.

[0051] The aim of this treatment is to improve the dryness of these sludges.

[0052] In all that follows, dryness refers to the mass percentage of dry matter in the corresponding medium.

[0053] The treatment carried out by the dehydration system 12 results in output products.

[0054] These output products mainly consist of dewatered sludge and centrate.

[0055] In what follows, "centrate" refers to the fraction of liquid removed from the sewage sludge entering the dewatering system. This centrate may contain suspended organic matter.

[0056] In the example shown in Figure 2, the dehydration system 12 includes a centrifuge, also called a high-pressure centrifuge. However, it should be understood that the invention is not limited to this embodiment of a dehydration system and may include any other system known in the prior art.

[0057] Furthermore, Figure 2 illustrates only one example of how to construct a centrifuge. Any other known example is also possible.

[0058] With reference to Figure 2, the centrifuge 12 has a body 20 extending along the longitudinal direction D and having a substantially symmetrical shape with respect to this direction D. In the example of Figure 2, the body 20 has a straight part and a narrowing part arranged along the longitudinal direction D. The straight part is formed, for example, by a straight circular cylinder and the narrowing part is formed by a truncated circular cone.

[0059] The body 20 defines an axial inlet 21 allowing the sewage sludge to be treated to be injected along this longitudinal direction D.

[0060] In its internal part, the body 20 includes a bowl 24 and a conveying screw 25 arranged one after the other along the longitudinal direction D.

[0061] In particular, the bowl 24 is arranged substantially in the right part of the body 20 and has a cylindrical body extending along the longitudinal direction D, one end of which opens internally into the axial inlet 21. The bowl 24 is arranged coaxially with the body 20 of the centrifuge 12 so that a retention space 26 is formed between the external walls of the bowl 24 and the internal walls of the body 20 of the centrifuge 12. Furthermore, the bowl 24 is configured to be rotated by a motor provided for this purpose at a configurable rotational speed.

[0062] The end of the bowl 24 opposite the end associated with the axial inlet 21 opens onto an internal part 27 of the body 20 of the centrifuge 12. This internal part 27 includes in particular the retention space 26, as defined above.

[0063] The conveying screw 25 is arranged coaxially with the bowl 24 while being spaced from it in the longitudinal direction D.

[0064] In addition, the conveying screw 25 is arranged in the narrowing part of the centrifuge body 20 12.

[0065] The conveyor screw 25 is configured to be rotated in the same way as the bowl 24, for example by a different motor than the one rotating the bowl 24. The rotation can be done at a configurable rotation speed.

[0066] The bowl 24 and the conveying screw 25 are therefore configured to be rotated at different rotational speeds.

[0067] Finally, the centrifuge 12 defines a transverse outlet 28 downstream of the conveying screw 25.

[0068] This transverse outlet 28, for example, is arranged in a direction perpendicular to the longitudinal direction D and allows, in particular, the recovery of the solid phase of the treated sewage sludge. In other words, this outlet 28 allows the recovery of dewatered sludge.

[0069] The body 20 of the centrifuge 12 further includes one or more outlets (not shown in Figure 2) for evacuating the centrate.

[0070] The body 20 of the centrifuge 12 also includes a polymer inlet (not shown in Figure 2) allowing polymers to be injected into the centrifuge 12 to improve the efficiency of separating the solid phase from the liquid phase of sewage sludge.

[0071] The centrifuge 12 or more generally the dehydration system 12 finally includes a control module 30 (visible in figure 1) allowing its operation to be controlled according to a plurality of operating parameters.

[0072] The operating parameters include, in particular, at least one of the elements chosen from the group comprising:

[0073] sludge flow rate;

[0074] polymer flow rate;

[0075] relative speed of rotation screw / bowl;

[0076] screw / bowl pair. In some embodiments, all the aforementioned elements are chosen to control the operation of the dehydration system 12. For example, the operating parameters are composed of these aforementioned elements.

[0077] In certain modes, the chosen operating parameters allow control of the operation of any type or model of the dehydration system 12.

[0078] In particular, the sludge flow type parameter allows control of the flow of sewage sludge injected into the centrifuge 12 via the axial inlet 21. To do this, the control module 30 is configured for example to control the operation of a pump associated with the axial inlet 21.

[0079] Similar to the sludge flow rate, the polymer flow rate parameter allows control of the polymer supply to the centrifuge 12. For this purpose, the control module 30 is configured for example to control the operation of a pump associated with the polymer inlet.

[0080] The screw / bowl relative speed parameter allows control of the relative rotational speed of these two components. For this purpose, the control module 30 is configured, for example, to control the rotational speed of each of the motors associated with the bowl 24 and the conveyor screw 25, or at least the ratio of their rotational speeds. In some examples, the control module 30 is configured to control the rotational speed of the conveyor screw 25 only.

[0081] Finally, the screw / bowl torque parameter signifies a relative torque setpoint between the conveying screw 25 and the bowl 24 during slurry rotation. The control module 30 is configured, for example, to maintain the torque according to this setpoint by measuring the actual torque inside the centrifuge 12 and adjusting the other operating parameters, particularly the relative screw / bowl rotation speed.

[0082] Returning to the description in Figure 1, the control system 10 comprises an input module 41, a processing module 42 and an output module 43.

[0083] According to one embodiment, each of these modules 41 to 43 is implemented at least partially in the form of software. In such a case, the control system 10 further includes a processor for implementing this software and memory for storing it, at least temporarily.

[0084] In addition or alternatively, at least some of these modules 41 to 43 are implemented at least partially using a programmable logic circuit such as an FPGA (Field-Programmable Gate Array).

[0085] The input module 41 allows the input data necessary for the operation of the control system 10 to be received. In particular, the input module 41 is connected to a characterization device 50 allowing the output products of the dehydration system 12 to be characterized by a plurality of characterization parameters, as will be explained in more detail later.

[0086] In other words, the input module 41 is configured to receive characterization parameters from this characterization device 50.

[0087] The input module 41 is also connected to a rule base 51 and a parameter base 52.

[0088] The parameter base 52 comprises a plurality of hyperparameters defined for each type / model of the dehydration system 12 that may be used. A set of hyperparameters associated with a particular type / model of the dehydration system 12 includes extrema of values ​​tolerated by that target dehydration system 12. These extrema include, for example, the minimum and maximum capacity of the dehydration system 12, the minimum and maximum torque, etc. Advantageously, each hyperparameter is associated with a particular type of operating parameter of the dehydration system 12 and presents, for example, an extremum of that parameter tolerated by the dehydration system 12.

[0089] Rule base 51 includes a plurality of fuzzy rules.

[0090] In particular, fuzzy rules are configured to link at least some characterization parameter values ​​generated by the characterization device 50 to actions consisting of modifying the operating parameters of the dehydration system 20.

[0091] Advantageously, each fuzzy rule is determined from the knowledge of experts familiar with the functioning of the dehydration system 12.

[0092] According to one embodiment, the fuzzy rules can be determined beforehand from information obtained from interviews with different experts in order to take into account the variability of practices relating to the use of the dehydration system 12. In some embodiments, each fuzzy rule can link the current state of dryness or the rate of capture of suspended matter in the output products to a number of actions which consist of modifying the operating parameters of the dehydration system 12.

[0093] Advantageously, each fuzzy rule is based on at least one input parameter independent of the dehydration system 12. In other words, such an input parameter is independent of the type and / or model of the dehydration system 12. Even more advantageously, each fuzzy rule is based solely on input parameters independent of the dehydration system 12. Generally, a fuzzy rule within the meaning of the present invention consists of several elements: propositions, expressions (unary, binary, etc.), and conclusions. The propositions may include linguistic variables within the framework of fuzzy logic. A rule comprises several expressions (premises, conclusions) and one or more operators (for example, standard predefined logical operators such as "if," "then," etc., as well as specific operators such as "before," "after," and "while").For example, a fuzzy rule might be "if the capture rate is low, then the torque reduction is moderate." An expression can include a sequence of characters (for example, "if the capture rate is low" and "the torque reduction is moderate"). An expression can be a regular, rational, or formal expression.

[0094] Advantageously, the premise of a fuzzy rule is defined by at least one parameter whose value may be uncertain. The conclusion of such a rule is defined by its membership function, and the modification of this function after inference depends, in particular, on the degree of uncertainty associated with the value of said parameter.

[0095] Figure 4 represents an example of the form in which the rule "if the capture rate is low, then the reduction in torque is moderate" can be stored in rule base 51.

[0096] According to the example in this figure, the rule comprises a premise P (i.e., the rule condition: "if the capture rate is low"), represented by the left-hand graph, and a conclusion C (i.e., "the torque reduction is moderate"), represented by the right-hand graph. The curves shown in these graphs are the membership functions of the fuzzy terms "low" for the variable "capture rate" and "moderate" for the output variable "torque reduction," respectively. In Figure 4, the capture rate is characterized by the parameter CR, ranging from 88% to 100%, the torque reduction value is characterized by the parameter TR, ranging from 0 to 10%, and the degree of membership is characterized by the parameter D, ranging from 0 to 1.

[0097] In the example shown, the membership functions are of broken line or trapezoid type, but they could be any form of membership functions (e.g., Gaussian, sigmoid, triangle, etc.).

[0098] Advantageously, a fuzzy rule according to the invention includes vocabulary belonging to the natural language (i.e., the words or sequences of characters are words of the natural language).

[0099] As previously stated, the characterization device 50 allows the output products from the dehydration system 12 to be characterized. In particular, the characterization device 50 allows the generation of characterization parameters relating to the output products of the dehydration system 12.

[0100] The characterization device 50 is configured, for example, to characterize these output products in real time or according to predetermined time intervals ranging from one minute to one or more hours, for example. In other words, these time intervals define a characterization frequency that can, for example, remain constant throughout the operation of the dehydration system 12.

[0101] The characterization parameters may vary depending on the composition of the output products.

[0102] In the above example, when the output products include dewatered sludge and centrate, the characterization parameters include at least one of the elements chosen from the group comprising:

[0103] the shade of centrât;

[0104] the flocculation of the centrate;

[0105] the dryness of the dehydrated sludge;

[0106] the capture rate.

[0107] The capture rate can form a derived parameter as explained below. According to different embodiments, the characterization device 50 can include sensors for generating the characterization parameters as defined above.

[0108] Advantageously, these sensors may have lower accuracy compared to those used, for example, in the prior art.

[0109] According to another embodiment, the number of these sensors is reduced compared to that usually used in the prior art.

[0110] In some embodiments, the characterization device 50 includes an interface for interaction with an operator in charge of the operation of the dehydration system 12.

[0111] For example, in such a case, the operator has the possibility of introducing into the characterization device 50 subjectively determined characterization parameters.

[0112] This can, for example, be done from a visual observation of the output products or from the use of imprecise measuring devices by the operator such as images of the output products or rulers or any other available measuring device.

[0113] According to some embodiments, the characterization device 50 allows the generation of characterization parameters by implementing artificial intelligence techniques based on previously acquired measurements (machine learning).

[0114] In particular, in such a case, the characterization device 50 can for example be configured to generate characterization parameters from, for example, images of output products or from any other easily measurable physical parameter.

[0115] Of course, all the aforementioned embodiments of the characterization device 50 can be combined with each other so that the characterization device 50 uses any technically possible combination of the aforementioned characterization parameter generation techniques.

[0116] The control device 10 enables the implementation of a process for controlling the operation of the dehydration system 12. This process will now be explained with reference to Figure 3, which presents a flowchart of its steps.

[0117] It is initially considered that rule base 51 is formed with fuzzy rules as explained previously.

[0118] It is also assumed that the dehydration system 12 is in operation using, for example, default control parameters.

[0119] During an initial step 110, the input module 41 acquires characterization parameters generated by the characterization device 50.

[0120] In particular, these characterization parameters were generated by this characterization device 50 to characterize the output products from the dehydration system 12, for example in real time or according to the predetermined frequency.

[0121] To generate these operating parameters, the characterization device 50 uses one of the techniques, as explained previously.

[0122] Then, at the end of this step 110, the input module 41 transmits the characterization parameters to the processing module 42.

[0123] The input module 41 also transmits at least some of the fuzzy rules from the rule base 51 to the processing module 42 and hyper-parameters from the parameter base 52 and associated with the dehydration system 12.

[0124] In a subsequent step 120, the processing module 42 then processes the data transmitted by the input module 41 in order to determine a control instruction applicable to the dehydration system 12.

[0125] To do this, according to an example implementation, the processing module 42 implements two sub-steps.

[0126] In a first sub-step 121, the processing module 42 performs a pre-processing of the characterization parameters as received by the input module 41. This pre-processing may include the determination of at least one derived parameter and / or the conversion of the characterization parameters.

[0127] The determination of at least one derived parameter is carried out, for example, according to business rules known as such.

[0128] This determination may include, for example, determining the capture rate of the centrate from the color and / or flocculation of the centrate.

[0129] In some modes, a derived parameter is determined directly by the characterization device 50 before being transmitted to the input module 41.

[0130] Thus, in the present description, the term "characterization parameter" is used interchangeably to designate a parameter determined directly from the measurements or indirectly (i.e. derived parameter) by the characterization device 50 or by the processing module 42.

[0131] During the conversion of characterization parameters, the processing module 42 converts the characterization parameters into input parameters as defined by the corresponding fuzzy rules, by combining them with hyperparameters associated with the dehydration system 12. In this case, the hyperparameters are, for example, the extreme values ​​of the ranges tolerated by the target dehydration system 12, such as the minimum and maximum target dryness levels, which allows the rules to be written based on their normalized values. Mathematically, the following formula is applied to the dryness example:

[0132] >

[0133]

[0134] The dryness characterization parameter me The certainty is converted into a normalized siccity input parameter via the hyperparameters minimum siccity and siccity ma ximum. In the rule base, the fuzzy rules only apply to the normalized dryness input, making them easily adaptable to several dehydration systems with different dryness targets. Without this conversion, the membership functions associated with dryness would have to be redefined for each dehydration system.12

[0135] In other words, according to a particular example of implementation of this substep 121, the processing module 42 normalizes the characterization parameters according to the type of dehydration system 12 used.

[0136] For example, such normalization might involve converting characterization parameters into extrema of those parameters usable by the dehydration system 12. According to this example, the processing module 42 could, for instance, determine a maximum and a minimum value for each type of characterization parameter, and then determine the corresponding input parameter value as the ratio between the difference between the current value of that parameter and its minimum value, and the difference between its maximum and minimum values. The corresponding hyperparameters (i.e., maximum and minimum values) could, for example, be predetermined, as explained previously.

[0137] Of course, any other form of normalization can be applied to make the characterization parameters conform to the input parameters of the fuzzy rules.

[0138] In a second substep 122, the processing module 42 applies fuzzy rules according to fuzzy logic to determine a control setpoint for the dehydration system.

[0139] For this, the processing module 42 implements, for example, a technique called Generalized Modus Ponens, which allows a decision to be made in the context of fuzzy logic.

[0140] According to fuzzy logic, processing module 42 infers the fuzzy rules independently, resulting in a transformation of the conclusion's membership function. Then, processing module 42 aggregates the outputs of all the rules. In other words, processing module 42 determines a single membership function for the conclusion's domain. Finally, processing module 42 implements a "defuzzification" technique that associates a resulting control instruction with this membership function.

[0141] Each control instruction may include one or more actions to be performed in relation to the operating parameters.

[0142] Such an action could, for example, include one of the elements chosen from the group comprising:

[0143] - increasing / decreasing the screw / bowl torque;

[0144] - increase / decrease in relative speed,

[0145] - increase / decrease in sludge flow rate;

[0146] - increase / decrease in polymer flow rate;

[0147] - No action.

[0148] Of course, other actions as well as actions comprising a combination of the aforementioned actions are also possible.

[0149] Figures 5 and 6 illustrate two examples of the application of the fuzzy rule as explained in relation to Figure 4.

[0150] In the example in Figure 5, the processing module 42 receives a precise / certain value for the capture rate CR. This value, which can come from a sensor or a laboratory analysis, indicates that the capture rate is 91% (left side of Figure 5). We therefore observe a membership function that is a Dirac delta function centered on the value of 91%. The application of the Generalized Modus Ponens technique by the processing module 42 will thus adapt the conclusion of the fuzzy rule. Here, the rule indicates that a value of 91% cannot be considered a completely low capture rate, and therefore the torque reduction should not be as moderate as suggested by the rule.

[0151] In the example in Figure 6, the processing module 42 receives an imprecise / uncertain value. This could be a measurement from an inaccurate sensor (for example, a high-speed sensor), or pre-established membership functions associated with a color chart and entered by the operator. Instead of a Dirac delta function on the left, the value displays a trapezoid, which can be interpreted as "the capture rate is approximately between 90% and 92%." The conclusion on the right is transformed differently from that in Figure 5, notably by giving more weight to the conclusion "the torque reduction is moderate" than before.

[0152] In other words, in the example of these figures, we see that the Generalized Modus Ponens technique allows us to adapt the degree of activation of a fuzzy rule according to the degree of adequacy of the value of the corresponding characterization parameter(s), i.e. the characterization parameter(s) on which this fuzzy rule depends.

[0153] Specifically, the degree of adequacy is the numerical measure, between 0 and 1, that expresses how closely an input value matches a fuzzy concept defined by the fuzzy rule. The degree of fuzzy rule activation is the numerical value, between 0 and 1, that quantifies the intensity with which that rule is applied.

[0154] At the end of substep 122, the processing module 42 transmits the determined control instruction to the output module 43.

[0155] In a subsequent step 130, the output module 43 applies this control instruction in order to control the operation of the dehydration system 12.

[0156] To do this, the output module 43 sends, for example, this control instruction to the control module 30 of this dehydration system 12.

[0157] This control module 30 then applies control parameters according to the received control instruction. This allows the dehydration system 12 to operate according to the new instruction.

[0158] The control process can then be repeated, for example, cyclically or after a predetermined interval (for example with a predetermined repetition frequency) or following an explicit command from the operator.

[0159] Of course, many other embodiments of the invention are possible.

Claims

DEMANDS 1. A method for controlling the operation of a sludge dewatering system (12), the dewatering system (12) being configured to treat sewage sludge by producing output products, the method comprising the following steps: - acquisition (110) of a plurality of characterization parameters characterizing output products; - determination (120) of a control instruction by applying fuzzy rules to the characterization parameters according to a fuzzy logic, the fuzzy rules being derived from a predetermined base of fuzzy rules (51), each fuzzy rule being based on at least one input parameter independent of the dehydration system (12); - application (130) of the control instruction to the dehydration system (12); wherein: - the determination step (120) includes a substep (121) of preprocessing the characterization parameters including the conversion of the characterization parameters into input parameters; - fuzzy rules link at least certain values ​​of characterization parameters to actions consisting of modifying operating parameters of the dehydration system (12); - The operating parameters include a sewage sludge flow rate and a polymer flow rate for treating the sewage sludge; and - the dehydration system (12) is a centrifuge comprising a bowl (24) and a conveying screw (25), and in which the operating parameters further include a relative speed of rotation screw / bowl and a screw / bowl torque.

2. A process according to claim 1, wherein the output products comprise dewatered sludge and a centrate, and wherein the characterization parameters comprise an estimation of the dryness of the dewatered sludge, an estimation of the color of the centrate and / or an estimation of the flocculation of the centrate.

3. A method according to claim 2, wherein the estimates of the color and / or flocculation and / or dryness of the output products are determined by comparing the color and / or flocculation and / or dryness of the output products with that of predetermined samples.

4. A method according to any one of the preceding claims, wherein the characterization parameters are determined by one or more low-precision sensors and / or by an operator.

5. A method according to any one of the preceding claims, wherein the control instruction comprises one or more actions determined by fuzzy rules.

6. A method according to any one of the preceding claims, wherein each fuzzy rule is determined from the knowledge of the dehydration system experts (12).

7. A method according to any one of the preceding claims, wherein the conversion of characterization parameters into input parameters is done using hyperparameters, each hyperparameter being associated with one of the characterization parameters and defined according to the dehydration system (12).

8. Method according to claim 7, wherein each hyperparameter has an extremum of the associated characterization parameter which is applicable to the dehydration system (12).

9. A method according to any one of the preceding claims, wherein the substep (121) of preprocessing the characterization parameters further comprises the determination of a parameter derived from at least some of the characterization parameters.

10. A method according to claim 9 taken in combination with claim 2, wherein said derived parameter is a substrate capture rate determined from estimates of centrate shade and / or centrate flocculation.

11. A method according to any one of the preceding claims, wherein the determination step further comprises a substep (122) of applying fuzzy rules according to fuzzy logic using a Generalized Modus Ponens technique.

12. A method according to claim 11, wherein the Generalized Modus Ponens technique is configured to adapt the degree of activation of a fuzzy rule in 18 function of the degree of adequacy of the value of the corresponding characterization parameter(s).

13. A method according to any one of the preceding claims, wherein the pretreatment of the characterization parameters includes a normalization of these characterization parameters according to the type of dehydration system (12) used.

14. Control system (10) for the operation of a sludge dewatering system (12), comprising technical means (41, 42, 43) configured to implement the process according to any one of the preceding claims.

Images