Arsenic concentration prediction method and device for water, electronic equipment and storage medium

Abstract

The application discloses an arsenic concentration prediction method and device of water, electronic equipment and storage medium, and relates to the field of water treatment. The method comprises the following steps: acquiring the arsenic concentration value, pH value, oxidation-reduction potential, flow and temperature parameters of water; determining the dosing amount of reagent according to the aforementioned parameters; taking the arsenic concentration value of water before treatment, the dosing amount of hypochlorite and iron salt as input samples, and taking the arsenic concentration value of water after treatment as output samples; pre-training an error inverse feedback neural network prediction model to generate an arsenic concentration prediction model; adjusting the size of the dosing amount of hypochlorite and / or iron salt, combining the arsenic concentration prediction model to obtain the predicted arsenic concentration value of water after treatment, correcting the arsenic concentration prediction model, and optimizing the size of the dosing amount of hypochlorite and iron salt. Through the method, the problem that current water arsenic removal relies on human experience, cannot be fully automated, and the inaccurate dosing amount of reagent easily leads to reagent waste is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of water treatment, in particular to an arsenic concentration prediction method and system for water, an electronic device and a storage medium. BACKGROUND

[0002] Arsenic concentration needs to be removed in the treatment process of municipal water and groundwater. Since there is a lack of real-time monitoring means, the calculation of the dosing amount is relatively rough. A fixed dosing amount is often set according to the design value, and a periodic sampling and testing method is used for adjustment. This method has the problem that the addition amount of the reagent cannot be easily adjusted in real time according to the arsenic concentration, resulting in waste of the reagent. SUMMARY

[0003] In view of the above problems, the present application provides an arsenic concentration prediction method and system for water, an electronic device and a storage medium, which solves the problem that the addition amount of the reagent cannot be easily adjusted in real time according to the arsenic concentration, resulting in waste of the reagent.

[0004] In order to solve the above problems:

[0005] In a first aspect, the present application provides an arsenic concentration prediction method for water, which comprises:

[0006] obtaining an arsenic concentration value, a pH value, an oxidation-reduction potential, a flow rate and a temperature parameter in water;

[0007] determining a dosing amount of a reagent for reducing the arsenic concentration in water according to the arsenic concentration value, the pH value, the oxidation-reduction potential, the flow rate and the temperature parameter, the reagent including hypochlorite and iron salt;

[0008] using the arsenic concentration value of the water before treatment, the dosing amount of the hypochlorite and the dosing amount of the iron salt in the historical data as input samples, and using the arsenic concentration value of the water after treatment as an output sample, pre-training an error inverse feedback neural network prediction model to generate an arsenic concentration prediction model;

[0009] adjusting the size of the dosing amount of the hypochlorite and / or the iron salt, and combining the arsenic concentration prediction model to obtain a predicted arsenic concentration value of the water after treatment, and correcting the arsenic concentration prediction model to optimize the size of the dosing amount of the hypochlorite and the iron salt.

[0010] Optionally, the step of determining the dosing amount of the reagent for reducing the arsenic concentration in water according to the arsenic concentration value, the pH value, the oxidation-reduction potential, the flow rate and the temperature parameter comprises:

[0011] the dosing amount D1 of the hypochlorite:

[0012] D1 = a1 * F * (C(As) + 0.0034) * [|0.71 - 0.03 * pH - A ORP| + (0.71 - 0.03 * pH - A ORP )];

[0013] The dosage D2 of the iron salt:

[0014] D2 = a2 * 3F * C(As);

[0015] Wherein, F is the water flow, C(As) is the arsenic concentration of water, pH is the pH of water, A ORP is the oxidation-reduction potential of water, a1 is the comprehensive adjustment coefficient of hypochlorite dosage, a2 is the comprehensive adjustment coefficient of iron salt dosage.

[0016] Optionally, the step of adjusting the size of the dosage of hypochlorite and / or iron salt to correct the arsenic concentration prediction model to optimize the size of the dosage of hypochlorite and iron salt includes:

[0017] If the predicted value of the arsenic concentration of water output by the arsenic concentration prediction model is greater than or equal to the upper limit value of the preset arsenic concentration range, it indicates that the water treatment is unqualified; at this time, a1 is increased by a set step ratio within a set range, and after each adjustment of a1, the predicted value of the arsenic concentration of treated water is recalculated until the predicted value of the arsenic concentration of treated water is less than the upper limit value of the preset arsenic concentration range, indicating that the water treatment is qualified, at which time the a1 value, a2 value, and the predicted dosage of sodium hypochlorite and iron salt are recorded and output.

[0018] If a1 is increased to the upper limit value of the set range, and the predicted value of the arsenic concentration of treated water is still greater than or equal to the upper limit value of the preset arsenic concentration range, a2 is increased by a set step ratio within a set range, and after each adjustment, the predicted value of the arsenic concentration of treated water is recalculated until the predicted value of the arsenic concentration of treated water is less than the upper limit value of the preset arsenic concentration range, the a1 value, a2 value, and the predicted dosage of sodium hypochlorite and iron salt are recorded and output.

[0019] Optionally, the step of adjusting the size of the dosage of hypochlorite and / or iron salt to correct the arsenic concentration prediction model to optimize the size of the dosage of hypochlorite and iron salt includes:

[0020] If a2 is increased by a set step ratio within a set range until it reaches the upper limit value of the set range, and the predicted value of the arsenic concentration of treated water is still greater than or equal to the upper limit value of the preset arsenic concentration range, an alarm is output.

[0021] Optionally, the step of adjusting the size of the dosage of hypochlorite and / or iron salt to correct the arsenic concentration prediction model to optimize the size of the dosage of hypochlorite and iron salt includes:

[0022] If the predicted value of the arsenic concentration of the treated water is less than the lower limit of the preset arsenic concentration range, record the dosage of the agent to be added, the value of a1 and the value of a2;

[0023] Synchronously decrease the value of a1 and the value of a2 by a preset step ratio within a preset range, each time of adjustment re-predicts the predicted value of the arsenic concentration of the treated water, until the predicted value of the arsenic concentration of the treated water is greater than or equal to the lower limit of the preset arsenic concentration range and less than the upper limit of the preset arsenic concentration range, record the value of a1 and the value of a2 at this time, and output the predicted dosages of sodium hypochlorite and iron salt at this time.

[0024] Optionally, periodically acquire the arsenic concentration value, the pH value, the oxidation-reduction potential, the flow rate and the temperature parameter of the water, and acquire the detected value of the arsenic concentration of the treated water according to a preset lag time; put the above parameters into the self-optimizing learning module of the error inverse feedback neural network prediction model, correct the weight parameter in the error inverse feedback neural network prediction model, and periodically update to the error inverse feedback neural network prediction model.

[0025] Optionally, the arsenic concentration value of the water is monitored by using an online detection device;

[0026] The pH value, the oxidation-reduction potential, the flow rate and the temperature parameter of the water are monitored by using an online monitoring or metering device.

[0027] In a second aspect, the present application provides a device for predicting the arsenic concentration of water, the device comprising:

[0028] A first module is adapted to acquire the arsenic concentration value, the pH value, the oxidation-reduction potential, the flow rate and the temperature parameter of the water;

[0029] A second module is adapted to determine the dosages of sodium hypochlorite and iron salt for reducing the arsenic concentration of the water according to the arsenic concentration value, the pH value, the oxidation-reduction potential, the flow rate and the temperature parameter;

[0030] A third module is adapted to use the arsenic concentration value of the water before treatment and the dosages of sodium hypochlorite and iron salt as input samples, and use the arsenic concentration value of the treated water as an output sample, to pre-train an error inverse feedback neural network prediction model to generate an arsenic concentration prediction model;

[0031] A fourth module is adapted to correct the arsenic concentration prediction model by adjusting the size of the dosages of sodium hypochlorite and / or iron salt, and combining the arsenic concentration prediction model to obtain the predicted value of the arsenic concentration of the treated water, to optimize the size of the dosages of sodium hypochlorite and iron salt.

[0032] In a third aspect, the present application provides an electronic device, comprising: a processor; and a memory arranged to store computer-executable instructions that, when executed, cause the processor to perform the method for predicting the arsenic concentration of water as described in any of the above.

[0033] In a fourth aspect, the present application provides a computer readable storage medium storing one or more programs, which when executed by a processor, implement the method for predicting arsenic concentration of water according to any one of the above aspects.

[0034] In summary, the technical solution of the present application achieves the following technical effects:

[0035] The method for predicting arsenic concentration of water provided by the present application determines the dosing amount of the medicament for reducing the arsenic concentration in water according to the arsenic concentration value, pH value, oxidation-reduction potential, flow rate and temperature parameters, and the medicament includes hypochlorite and iron salt; the arsenic concentration value of the water before treatment, the dosing amount of the hypochlorite and the iron salt in the historical data are taken as input samples, and the arsenic concentration value of the water after treatment is taken as an output sample, and the error inverse feedback neural network prediction model is pre-trained to generate an arsenic concentration prediction model; then the size of the dosing amount of the hypochlorite and / or the iron salt is adjusted, the arsenic concentration prediction value of the water after treatment is obtained by combining the arsenic concentration prediction model, and the arsenic concentration prediction model is corrected to optimize the size of the dosing amount of the hypochlorite and the iron salt; by this method, the problem that the current water arsenic removal relies on human experience, cannot be fully automated, the medicament dosing feedback is lagging, and the inaccurate medicament dosing amount easily leads to medicament waste is solved. BRIEF DESCRIPTION OF DRAWINGS

[0036] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are intended to further assist in understanding the preferred embodiments, and are not intended to limit the application. Moreover, like reference numerals designate like parts throughout the several views in the drawings. In the drawings:

[0037] Figure 1 A flowchart of the method for predicting arsenic concentration of water provided by the present application is shown;

[0038] Figure 2 A flowchart of step 140 according to an embodiment of the present application is shown;

[0039] Figure 3 A flowchart of the method for removing arsenic from water in a water treatment plant according to an embodiment of the present application is shown;

[0040] Figure 4 A structural diagram of the device for predicting arsenic concentration of water according to an embodiment of the present application is shown;

[0041] Figure 5 A structural diagram of an electronic device according to an embodiment of the present application is shown;

[0042] Figure 6A structural diagram of a computer readable storage medium according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0043] Exemplary embodiments of the present application will be described in greater detail below with reference to the accompanying drawings. Although exemplary embodiments of the present application are shown in the drawings, it is understood that the present application can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided so that the present application can be more thoroughly understood, and so that the scope of the present application can be accurately conveyed to those skilled in the art.

[0044] Figure 1 A method for predicting arsenic concentration of water according to an embodiment of the present application is shown, as shown in Figure 1 The prediction method comprises:

[0045] Step 110: Obtain the arsenic concentration value, pH value, oxidation-reduction potential, flow rate and temperature parameters of the water;

[0046] In this step, the arsenic concentration value of the water is monitored by using an online detection device; the pH value, oxidation-reduction potential, flow rate and temperature parameters of the water are monitored by using an online monitoring or metering device.

[0047] Step 120: Determine the dosing amount of a medicament for reducing the arsenic concentration of the water according to the arsenic concentration value, pH value, oxidation-reduction potential, flow rate and temperature parameters; wherein the medicament comprises a hypochlorite and an iron salt;

[0048] In this step, the addition amount of the hypochlorite and the iron salt can be determined according to the experience of workers, or can be obtained by a scientific calculation method.

[0049] Step 130: Use the arsenic concentration value of the water before treatment, the dosing amount of the hypochlorite and the iron salt in the historical data as input samples, and use the arsenic concentration value of the water after treatment as an output sample, to pre-train an error inverse feedback neural network prediction model to generate an arsenic concentration prediction model;

[0050] The error inverse feedback neural network in the present embodiment is also called a BP neural network, which is a neural network that propagates errors in the backward direction. It comprises an input layer, a hidden layer and an output layer. The basic idea is the gradient descent method. There is a certain mathematical relationship between the error of the intermediate hidden layer and the error of the last layer. Just like the error is propagated back, so it is a BP neural network. In this step, however, the neural network is used after training with historical data to predict whether the arsenic concentration of the water after treatment is within a preset arsenic concentration range (i.e., meets the water quality requirements).

[0051] It should be noted that the trained arsenic concentration prediction model can output the arsenic concentration prediction value of the treated water according to the arsenic concentration value of the untreated water, the dosing amount of the hypochlorite and the iron salt, and then it can be judged whether the arsenic concentration of the treated water is qualified; however, in actual application, when the arsenic concentration of the treated water is in the qualified condition, how to control the dosing amount of the medicament to be optimal to save the use amount of the medicament is a problem to be solved by the embodiments of the present application, and for this, the method further includes the following steps:

[0052] Step 140: The arsenic concentration prediction model is corrected by adjusting the dosing amount of the hypochlorite and / or the iron salt, and combining the arsenic concentration prediction value of the treated water obtained by the arsenic concentration prediction model, so as to optimize the dosing amount of the hypochlorite and the iron salt, and then the dosing amount of the medicament (the hypochlorite and the iron salt) is controlled to be minimum when the arsenic concentration prediction value of the treated water is in the preset arsenic concentration range.

[0053] In combination with the above, through the arsenic concentration prediction method of water provided by the embodiments, the problems of current water arsenic removal depending on human experience, being unable to be fully automated, medicament dosing feedback lag, and inaccurate medicament dosing amount leading to medicament waste are solved, the automation of arsenic concentration prediction and treatment of water is realized, and the precision and accuracy of arsenic concentration prediction and treatment of water are improved.

[0054] In one embodiment, in step 120, the calculation method of the dosing amount of the medicament for reducing the arsenic concentration in water includes:

[0055] The dosing amount D1 of the hypochlorite is:

[0056] D1 = α1 * F * (C(As) + 0.0034) * [|0.71-0.03*pH-A ORP + (0.71-0.03*pH-A ORP )];

[0057] The dosing amount D2 of the iron salt is:

[0058] D2 = α2 * 3F * C(As);

[0059] Wherein, F is the water flow, C(As) is the arsenic concentration of water, pH is the pH of water, A ORP is the oxidation-reduction potential of water, α1 is the comprehensive adjustment coefficient of the dosing amount of the hypochlorite, and α2 is the comprehensive adjustment coefficient of the dosing amount of the iron salt.

[0060] It should be noted that the principle of arsenic removal in the embodiments of the present application is:

[0061] Principle 1: Adopting adding iron salt to remove arsenic, the principle is to promote the transformation of dissolved arsenic in water to insoluble arsenic-containing products and further separation. For natural water, the main process is co-precipitation and adsorption. The arsenic removal rate is related to the initial arsenic content in water and the iron-arsenic ratio in water, therefore, the formula of iron salt dosage D2 is based on this linear relationship.

[0062] Principle 2: Arsenic in natural water often presents +3 valence state [arsenite, As(III)] and +5 valence state [arsenate, As(V)], the ratio of the two valence states is mainly affected by pH, and there is a certain dominant regional distribution relationship. Studies have shown that the removal effect of iron salt on As(V) is better than that on As(III), so the water to be treated is often pre-oxidized during the arsenic removal process. Therefore, the hypochlorite used in this method is an oxidizing agent, and the formula of its dosage D1 is: the dosage is related to the arsenic content that needs to be oxidized, and also related to the lack of oxidation, and this oxidation is related to pH and ORP. The temperature parameter is not an independent variable of the formula, but only one of the influencing factors of the adjustment coefficient.

[0063] For the dosage of hypochlorite D1:

[0064] D1=α1*F*(C(As)+0.0034)*[|0.71-0.03*pH-A ORP |+(0.71-0.03*pH-A ORP )];

[0065] In the formula, 0.71 is an empirical value slightly higher than the reaction potential difference between AsO4 3- and AsO3 3- , -0.03*pH is a simplification of the influence of pH on the oxidation-reduction potential in the solution according to the Nernst equation (since the target is to approach complete oxidation, the lnAs 3- / As 5- value in the Nernst equation is too small to be ignored), A ORP is the original ORP value of the solution;

[0066] [|0.71-0.03*pH-A ORP |+(0.71-0.03*pH-A ORP )] is a shielding method, when the oxidation in the solution is sufficient (the oxidation of the solution after pH correction is higher than the demand of arsenic oxidation), this item is 0, no additional oxidizing agent is needed; when the oxidation in the solution is insufficient, the amount of added agent is proportional to the lack of oxidation, this item reflects the lack of oxidation in the solution when oxidizing trivalent arsenic as the target, measured by oxidation-reduction potential.

[0067] C(As) is the arsenic concentration of water, here the arsenic concentration is used instead of the trivalent arsenic concentration because of the limitations of online detection means, it is difficult to directly measure the trivalent arsenic, so the total arsenic concentration is directly used for calculation, although it will cause excessive dosing, but its influence can be corrected by a1, 0.0034 is to maintain about 0.25mg / L of sodium hypochlorite in the produced water to maintain the additional additional dosage of residual oxidizing property.

[0068] In the formula, the product relationship between a1 correction factor and three items is because it is known that the sodium hypochlorite dosage is proportional to the flow rate, and is proportional to the sum of the trivalent arsenic concentration and the required residual oxidizing property, while the missing oxidizing property in the solution can only be determined to be positively correlated, and cannot be determined to be proportional, this algorithm defaults to a proportional relationship for simplification.

[0069] For the dosage D2 of the iron salt:

[0070] D2=α2*3F*C(As)

[0071] The coefficient 3 is the reaction coefficient of the reaction to form pentavalent arsenic insoluble matter, and a2 is the comprehensive adjustment coefficient of the iron salt dosage considering detection error, reaction completeness, reagent purity, temperature influence, pH influence, and correction factor.

[0072] In one embodiment, please refer to Figure 2 In step 140, the step of adjusting the size of the sodium hypochlorite and / or iron salt dosage to optimize the size of the sodium hypochlorite and iron salt dosage includes:

[0073] If the predicted arsenic concentration of water output by the arsenic concentration prediction model is greater than or equal to the upper limit value of the preset arsenic concentration range, it indicates that the water treatment is unqualified; at this time, a1 is increased by a set step ratio within a set range, and the predicted arsenic concentration of water after treatment is re-predicted after each adjustment of a1, until the predicted arsenic concentration of water after treatment is less than the upper limit value of the preset arsenic concentration range, which indicates that the water treatment is qualified, at this time, the a1 value, a2 value, and the predicted sodium hypochlorite and iron salt dosage are recorded and output.

[0074] If the predicted arsenic concentration of water after treatment is still greater than or equal to the upper limit value of the preset arsenic concentration range when a1 is increased to the upper limit value of the set range, a2 is increased by a set step ratio within a set range, and the predicted arsenic concentration of water after treatment is re-predicted after each adjustment, until the predicted arsenic concentration of water after treatment is less than the upper limit value of the preset arsenic concentration range, the a1 value, a2 value, and the predicted sodium hypochlorite and iron salt dosage are recorded and output.

[0075] If the arsenic concentration of the treated water is greater than or equal to the upper limit of the preset arsenic concentration range, the method further comprises the following steps of: outputting a dosing alarm.

[0076] If the arsenic concentration of the treated water is less than the lower limit of the preset arsenic concentration range, the method further comprises the following steps of: recording the dose of the dosing agent, the value of a1 and the value of a2.

[0077] The values of a1 and a2 are simultaneously reduced by a preset step ratio within a preset range, and the predicted arsenic concentration of the treated water is re-predicted each time the values of a1 and a2 are adjusted until the predicted arsenic concentration of the treated water is greater than or equal to the lower limit of the preset arsenic concentration range and less than the upper limit of the preset arsenic concentration range, the values of a1 and a2 at this time are recorded, and the predicted dose of sodium hypochlorite and iron salt at this time is outputted.

[0078] In one embodiment, the error inverse feedback neural network prediction model comprises at least two hidden layers and a sigmoid activation function, wherein the number of nodes of the hidden layers, the number of partitions and the learning rate are determined according to the input parameters, the output parameters and the training conditions.

[0079] In one embodiment, the input items are the arsenic concentration of the water and the calculated doses of sodium hypochlorite and iron salt based on the arsenic concentration of the water, the pH value, the oxidation-reduction potential, the flow rate and the temperature, the output item is the predicted arsenic concentration of the treated water, the first layer of the hidden layer has 14 nodes, the second layer has 11 nodes, and the learning rate n of the two layers is set to 0.1. The model is operated in an offline training manner using historical data and is regularly updated.

[0080] After the method is applied for a period of time, the arsenic removal effect of the concentrated brine reaches 99% based on the set target, and the cost of dosing agent is saved by 12%.

[0081] Please refer to Figure 3 , the following describes the specific steps of applying the method to the arsenic removal pretreatment of the influent of a water treatment plant, which comprises:

[0082] Step 210, obtaining the online monitoring arsenic concentration C(As) of the influent of the water treatment plant and the pH value, the oxidation-reduction potential A ORP , the flow rate F and the temperature T of the influent;

[0083] Step 220, the water enters an arsenic removal dosing reactor, and the doses of sodium hypochlorite and iron salt are calculated and determined according to the online monitoring arsenic concentration C(As) and the pH value, the oxidation-reduction potential A ORP , the flow rate F of the influent;

[0084] Step 230, taking the arsenic concentration value of the water before treatment, the dosing amount of the hypochlorite and the iron salt in the historical data as the input sample, and taking the arsenic concentration value of the water after treatment as the output sample, pre-training the error inverse feedback neural network prediction model to generate an arsenic concentration prediction model;

[0085] Step 240, adjusting the size of the dosing amount of the hypochlorite and / or the iron salt, and combining the arsenic concentration prediction model to obtain the predicted arsenic concentration value of the water after treatment, correcting the arsenic concentration prediction model to optimize the size of the dosing amount of the hypochlorite and the iron salt;

[0086] Step 250, using a high-precision online arsenic concentration monitor to obtain the arsenic concentration value of the water after arsenic removal, and feeding back the monitoring results to the error inverse feedback neural network method according to the set lag corresponding relationship for model self-optimization.

[0087] Specifically, in step 210, the online arsenic concentration monitor, the flow meter are regularly calibrated and maintained, and the oxidation-reduction potential and pH probes are regularly calibrated or replaced.

[0088] In step 240, specifically comprising:

[0089] Using the arsenic concentration prediction model to obtain the predicted arsenic concentration value of the water after treatment, if the predicted arsenic concentration value of the water after treatment is greater than or equal to 0.01 mg / L, it indicates that the water treatment is unqualified, at this time, α1 is increased by a set step ratio of 1% within a set range, and the predicted arsenic concentration value of the water after treatment is re-predicted each time, until the predicted arsenic concentration value of the water after treatment is less than 0.01 mg / L, which indicates that the water treatment is qualified, at this time, the α1 value, the α2 value, and the predicted dosing amount of the sodium hypochlorite and the iron salt are recorded and outputted;

[0090] When α1 is increased, until the predicted arsenic concentration value of the water after treatment is still greater than or equal to 0.01 mg / L when the initial α1 value is 400%, then α2 is increased by a set step ratio of 1%, and the predicted arsenic concentration value of the water after treatment is re-predicted each time, until the predicted arsenic concentration value of the water after treatment is less than 0.01 mg / L, and then the α1 value, the α2 value, and the predicted dosing amount of the sodium hypochlorite and the iron salt are recorded and outputted;

[0091] If α2 is increased until the predicted arsenic concentration value of the water after treatment is still greater than or equal to 0.01 mg / L when the initial α2 value is 200%, a dosing alarm is outputted;

[0092] If the predicted value of the arsenic concentration of the treated water is less than or equal to 0.001 mg / L, it indicates that the water treatment is qualified and meets the standard, but the dosage of the reagent may be too large. At this time, the dosage of the reagent to be added and the values of a1 and a2 are recorded, and the values of a1 and a2 are simultaneously reduced by a set step ratio within a set range. Each adjustment re-predicts the predicted value of the arsenic concentration of the treated water until the predicted value of the arsenic concentration of the treated water is greater than 0.001 mg / L and less than 0.01 mg / L at a certain value of a1 and a2. The values of a1 and a2 at this time are recorded, and the predicted dosages of sodium hypochlorite and iron salt at this time are output.

[0093] It should be noted that the qualified range of the arsenic concentration of the water in the present embodiment is between 0.001 mg / L and 0.01 mg / L.

[0094] In step 250, the arsenic concentration of the treated water is obtained by using a high-precision online arsenic concentration monitor. The monitoring results are fed back to the error inverse feedback neural network method according to a set hysteresis corresponding relationship, and the model is self-optimized. The neural network algorithm of error back propagation is used to optimize the model formula. The activation function uses sigmoid function, the input items are C(As), pH, Aorp, F, T, D1 and D2, and the output item is the arsenic concentration of the treated water. The learning rate n of the two hidden layers is set to 0.1. The model is run in the mode of offline training using historical data and regular updating.

[0095] Please refer to Figure 4 The present application also provides a water arsenic concentration prediction device 500, which comprises:

[0096] A first module 510 is adapted to obtain the arsenic concentration, pH value, oxidation-reduction potential, flow rate and temperature parameters of the water;

[0097] A second module 520 is adapted to determine the dosages of sodium hypochlorite and iron salt for reducing the arsenic concentration of the water according to the arsenic concentration, pH value, oxidation-reduction potential, flow rate and temperature parameters;

[0098] A third module 530 is adapted to use the arsenic concentration of the water before treatment and the dosages of sodium hypochlorite and iron salt in the historical data as input samples, and use the arsenic concentration of the treated water as output samples, to pre-train the error inverse feedback neural network prediction model to generate an arsenic concentration prediction model;

[0099] A fourth module 540 is adapted to adjust the dosages of sodium hypochlorite and / or iron salt, and combine the arsenic concentration prediction model to obtain the predicted value of the arsenic concentration of the treated water, and correct the arsenic concentration prediction model to optimize the dosages of sodium hypochlorite and iron salt.

[0100] Those skilled in the art will understand that the modules in the devices in the embodiments can be changed adaptively and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments can be combined into one module or unit or component, and furthermore can be divided into multiple sub-modules or sub-units or sub-components. Except that at least some of such features and / or processes or units are mutually exclusive, all combinations of all features disclosed in this specification (including accompanying claims, abstract and drawings) and all processes or units of any methods or apparatuses disclosed thus can be adopted. Unless explicitly stated otherwise, each feature disclosed in this specification (including accompanying claims, abstract and drawings) can be replaced by an alternative feature providing the same, equivalent or similar purpose.

[0101] The various component embodiments of the present application can be implemented in hardware, or as software modules running on one or more processors, or in combination thereof. Those skilled in the art will appreciate that a microprocessor or a digital signal processor (DSP) can be used in practice to implement some or all of the functions of some or all of the components of the water quality early warning system according to the embodiments of the present application. The present application can also be implemented as a device or system program (e.g., a computer program and a computer program product) for performing part or all of the methods described herein. Such an application program implementing the present application can be stored in a computer readable medium, or can have one or more signals. Such signals can be downloaded from an Internet website, or provided on a carrier signal, or in any other form.

[0102] For example, Figure 5 A structural schematic of an electronic device according to one embodiment of the present application is shown. The electronic device 300 includes a processor 310 and a memory 320 arranged to store computer executable instructions (computer readable program code). The memory 320 can be an electronic memory such as a flash memory, an EEPROM (electrically erasable programmable read-only memory), an EPROM, a hard disk, or a ROM. The memory 320 has a storage space 330 storing computer readable program code 331 for performing any of the method steps in the above methods. For example, the storage space 330 for storing computer readable program code can include individual computer readable program codes 331 for implementing various steps in the above methods, respectively. The computer readable program code 331 can be read from or written to one or more computer program products. These computer program products include program code carriers such as a hard disk, a compact disk (CD), a memory card, or a floppy disk. Such computer program products are typically computer-readable storage media that are tangibly-embodied. For example, Figure 6 The computer readable storage medium described above. Figure 6A structural diagram of a computer readable storage medium according to one embodiment of the present application is shown. The computer readable storage medium 400 stores computer readable program code 331 for performing the method steps according to the present application, which can be read by the processor 310 of the electronic device 300, and when the computer readable program code 331 is run by the electronic device 300, causes the electronic device 300 to perform the steps of the methods described above, in particular, the computer readable program code 331 stored by the computer readable storage medium can perform the methods shown in any of the embodiments described above. The computer readable program code 331 can be compressed in a suitable form.

[0103] It should be noted that the above-mentioned embodiments illustrate rather than limit the application, and that those skilled in the art will be able to design many alternative embodiments without departing from the scope of the appended claims.

Claims

1. A method for predicting arsenic concentration in water, characterized in that, The method includes: Obtain arsenic concentration, pH value, redox potential, flow rate, and temperature parameters in the water; The dosage of the agent used to reduce the arsenic concentration in the water is determined based on the arsenic concentration value, pH value, redox potential, flow rate and temperature parameters. The agent includes hypochlorite and iron salt. Using the arsenic concentration, hypochlorite and iron salt dosage of the water before treatment from historical data as input samples, and the arsenic concentration of the water after treatment as output samples, the error inverse feedback neural network prediction model is pre-trained to generate an arsenic concentration prediction model. By adjusting the dosage of hypochlorite and / or iron salts and combining it with the arsenic concentration prediction model to obtain the predicted arsenic concentration of the treated water, the arsenic concentration prediction model is modified to optimize the dosage of hypochlorite and iron salts. The steps for determining the dosage of the reagent used to reduce the arsenic concentration in water based on the arsenic concentration, pH value, redox potential, flow rate, and temperature parameters include: Hypochlorite dosage D1: D1=α1*F*(C(As)+0.0034)*[|0.71-0.03*pH-A ORP |+(0.71-0.03*pH-A ORP )]; Iron salt dosage D2: D2 = α2 * 3F * C(As); Where F is the water flow rate, C(As) is the arsenic concentration in the water, pH is the pH of the water, and A ORP α1 is the redox potential of water, α2 is the comprehensive adjustment coefficient for hypochlorite addition, and α3 is the comprehensive adjustment coefficient for iron salt addition. The step of modifying the arsenic concentration prediction model by adjusting the dosage of hypochlorite and / or iron salt to optimize the dosage of hypochlorite and iron salt includes: If the arsenic concentration prediction model outputs a predicted arsenic concentration value for the water that is greater than or equal to the upper limit of the preset arsenic concentration range, it indicates that the water treatment is unqualified. At this time, α1 is increased by a set step ratio within the set range, and the predicted arsenic concentration value of the treated water is re-predicted after each adjustment of α1 until the predicted arsenic concentration value of the treated water is less than the upper limit of the preset arsenic concentration range, which indicates that the water treatment is qualified. At this time, the α1 value, α2 value, and predicted sodium hypochlorite and iron salt dosage are recorded and output. If the predicted arsenic concentration of the treated water is still greater than or equal to the upper limit of the preset arsenic concentration range when α1 is increased to the upper limit of the set range, α2 is increased by the set step ratio within the set range, and the predicted arsenic concentration of the treated water is re-predicted with each adjustment until the predicted arsenic concentration of the treated water is less than the upper limit of the preset arsenic concentration range. The values ​​of α1, α2, and the predicted dosage of sodium hypochlorite and iron salt are recorded and output. The arsenic concentration, pH value, redox potential, flow rate and temperature parameters in the water are acquired periodically, and the arsenic concentration of the treated water is detected according to the set lag time. The above parameters are substituted into the self-optimization learning module of the error inverse feedback neural network prediction model to correct the weight parameters in the error inverse feedback neural network prediction model, and are periodically updated to the error inverse feedback neural network prediction model.

2. The method according to claim 1, characterized in that, The step of modifying the arsenic concentration prediction model by adjusting the dosage of hypochlorite and / or iron salt to optimize the dosage of hypochlorite and iron salt further includes: If α2 is increased at a set step rate within the set range, and the predicted arsenic concentration of the treated water is still greater than or equal to the upper limit of the preset arsenic concentration range when the upper limit of the set range is reached, a dosing alarm will be output.

3. The method according to claim 1, characterized in that, The step of modifying the arsenic concentration prediction model by adjusting the dosage of hypochlorite and / or iron salt to optimize the dosage of hypochlorite and iron salt further includes: If the predicted arsenic concentration in the treated water is less than the lower limit of the preset arsenic concentration range, record the proposed dosage, α1 value, and α2 value. Within the set range, the values ​​of α1 and α2 are reduced synchronously at a set step ratio. Each adjustment re-predicts the arsenic concentration of the treated water until the predicted arsenic concentration of the treated water is greater than or equal to the lower limit of the preset arsenic concentration range and less than the upper limit of the preset arsenic concentration range. The values ​​of α1 and α2 at this time are recorded, and the predicted dosage of sodium hypochlorite and iron salt at this time is output.

4. The method according to claim 1, characterized in that, The arsenic concentration in the water was obtained by monitoring with online detection equipment; The pH value, oxidation-reduction potential, flow rate, and temperature parameters of water are obtained by using online monitoring or metering equipment.

5. A device for predicting arsenic concentration in water, characterized in that, The apparatus used in the method for predicting arsenic concentration in water according to claim 1 includes: The first module is suitable for obtaining arsenic concentration, pH value, redox potential, flow rate and temperature parameters in water; The second module is adapted to determine the dosage of hypochlorite and iron salts to reduce the arsenic concentration in water based on the arsenic concentration value, pH value, redox potential, flow rate, and temperature parameters. The third module is suitable for using the arsenic concentration value of the water before treatment, the dosage of hypochlorite and iron salt in historical data as input samples, and the arsenic concentration value of the water after treatment as output samples to pre-train the error inverse feedback neural network prediction model to generate an arsenic concentration prediction model. The fourth module is adapted to adjust the dosage of hypochlorite and / or iron salts, and to modify the arsenic concentration prediction model by combining the predicted arsenic concentration of the treated water with the arsenic concentration prediction model, so as to optimize the dosage of hypochlorite and iron salts.

6. An electronic device, characterized in that, The electronic device includes: a processor; and a memory arranged to store computer-executable instructions, which, when executed, cause the processor to perform the method for predicting the arsenic concentration in water according to any one of claims 1-4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores one or more programs that, when executed by a processor, implement the method for predicting the arsenic concentration in water according to any one of claims 1-4.

Images