Method for determining the specific power requirement of a belt conveyor system for bulk materials in operation under variable loading 5 The invention relates to a method for determining the specific power requirement of a belt conveyor system for bulk materials in operation under variable loading. In the simplest case, belt conveyor systems comprise an 10 elastomeric belt that is deflected around two drums and circulates endlessly. At least one of the two drums is electrically driven and drives the thus circulating -belt. Between the drums, which are generally denoted as head station and rear station, the belt is supported by 15 rotatable support rollers that are taken up by support roller frames. The support rollers serve t6 support and align the belt between the deflecting stations, and moreover determine the troughing of the belt. As already mentioned, the energy required to drive the 20 belt is introduced into the drums of the belt conveyor by electric drives. The power requirement of the belt conveyor is, on the one hand, dependent on the mass flow fed to the belt conveyor, while on the other hand it is dependent on the different movement resistances 25 of the belt conveyor, which exhibit a different appearance as a function of the state of alignment of the belt, of the contamination state of the latter, or of the state of maintenance of the rollers, etc. 30 It is customary to calculate the theoretical power requirement of a belt system in order to design the latter. According to this process, the different movement resistances (primary, secondary, climbing and special resistances) are determined individually. The 35 power requirement is calculated from the sum of these movement resistances multiplied by the belt speed. The largest portion of the movement resistance in a belt system belongs to the primary resistance, which is determined in a way similar to the determination of the 40 frictional resistance of a block pulled on an inclined - 2 plane. In this case, the specific movement resistance used hereby corresponds to a fictional coefficient of friction that can vary in dependence on the state of alignment of the belt system, diameter of the support 5 rollers, belt pretensioning, etc. This fictional coefficient of friction is usually estimated in order to determine the primary resistance of the belt system. During operation of the belt system, this specific value can undergo creeping deterioration, for example 10 owing to increasing contamination in the lower run but, in the case of shiftable belt systems, chiefly owing to change in the state of alignment of the belt supporting frames (underscouring by rain, displacement owing to inattentiveness during cleaning of the lower run by 15 means of an auxiliary unit). For this reason, the state of alignment and thus the specific energy consumption of the belt systems are frequently higher than necessary and cause substantial, avoidable energy costs. 20 It could now be assumed that the specific movement resistance of a belt system can be determined empirically simply by measuring power at a belt system that is operating. This would, for example, be 25 desirable in order to be able to draw conclusions therefrom in relation to the state of maintenance of the belt system, and/or in order to be able to check the effectiveness and the influence of the one or other maintenance measure with regard to the power 30 consumption of the belt system. However, in practice, it has proved to be difficult to specify a specific movement resistance or fictional coefficient of friction, since the mass flow fed to the conveying belt is not constant, as a rule, but fluctuates. This means 35 that the belt has different load cross sections over its length. It is possible to measure only the drive power of the motors as well as the current mass flow via a belt weigher possibly provided. In order to be able to draw conclusions in relation to the state of -3 maintenance of the belt system, the mass flow would need to be eliminated from consideration, it having emerged in experiments that it is impossible to derive any unique functional relationship between the drive 5 power and the belt load, and so the specification of a fictional coefficient of friction or specific movement resistance is subject to great uncertainty. It is therefore the object of the invention to provide 10 a method of the type mentioned in the beginning for the purpose of determining the specific power requirement of a belt conveyor system for bulk materials in operation under variable loading that is so reliable that it permits conclusions to be drawn in relation to 15 the state of the belt system. According to the invention, a method for determining the specific power requirement of a belt conveyor system for bulk materials in operation under variable 20 loading is proposed that comprises the following steps: - continuous determination of a segment related loading of the belt conveyor for n segments of the 25 same length; - calculating a load dependent movement resistance for each of the n segments, initially with the aid of an estimated specific movement resistance; - calculating power required per segment; 30 - calculating the total power requirement of the belt conveyor system; and determining a specific movement resistance by comparing the calculated total power requirement with a measured electric power requirement. 35 The inventive method proceeds in this case from the finding that different loading distributions of the belt are possible in principle in conjunction with the same average loading of a belt system. Depending on the - 4 load distribution, the power requirement of the belt system can be different although the mean value for the average load is the same. This gives rise in principle to the difficulty of deriving a functional relationship 5 between drive power and belt load. According to the invention, account is taken of this by determining different segment related loadings for n segments of the belt, that are of the same length. For 10 each of the n segments, this results in different primary movement resistances that are also dependent on the gradient of the belt profile in the relevant segment referred to a horizontal. It is possible in this way to determine a total power requirement of the 15 belt conveyor system that does justice to the different load cross section over the length of the belt. By continuously comparing the calculated theoretical total power requirement with a measured electric power requirement, it is possible to determine iteratively 20 the specific movement resistance on the basis of which conclusions can be drawn directly in relation to the state of maintenance of the belt system. The iterative determination of the specific movement 25 resistance is preferably performed by minimizing the spring squares sum from the comparison between calculated and measured power. A specific movement resistance which directly supplies information relating to the state of the belt system with regard to power 30 consumption can in this way be determined individually for each belt system in operation. Thus, the effectiveness of specific maintenance measures or other changes relevant with regard to the power consumption of the system can be directly determined during 35 operation of the belt system and be identified by being eliminated from the influences of uneven loading. With particular preference, the segment related loading is determined by means of a belt weigher during the -5 belt circulation, the clock cycle of the mass flow measurement being synchronized with the belt speed, such that the temporally varied loading can be assigned to the segments of the belt system, which differ with 5 reference to their power requirement. In order to be able to assign the measured mass flows to the individual segments of the belt system, the lengths of said segments are adapted to the clock cycle of the mass flow measurement. For example, if a belt weigher 10 supplies every two seconds a measured value that should constitute the mean value of the load over two seconds, and if the belt speed is 7.5 m/s, the belt system should be subdivided into segments 15 m long, and in each case the average gradient of the belt profile of 15 this segment should be determined for the purpose of calculating the movement resistance of the relevant segment. Expediently, the respective theoretical power 20 requirement is determined in each clock cycle At for all n segments of the belt system, and a theoretical total power requirement is determined in each clock cycle from the theoretical power requirement values determined in segment related fashion. 25 The minimization of the errors square sum of the deviations between theoretical and measured power requirements over all time increments is expediently performed over a time period that is representative of 30 all occurring load fluctuations. The invention will be explained below with the aid of an example and with reference to the attached drawing of the figure. The figure illustrates in a relatively 35 simple schematic way a belt conveyor system having two different load distributions in conjunction with the same average loading of a belt system with four segments. By way of example for a belt conveyor as applied in modern open cast brown coal mines, the belt - 6 width is approximately 2800 mm. The mass flow or flow of bulk materials can fluctuate between 0 and 35000 t/h. 5 As already mentioned initially, the figure illustrates two different loading states of a belt system. It is evident that - despite the same average loading - the power requirement in the lower case is substantially higher than in the upper case, specifically 10 particularly because the profile of the belt system is not horizontal, and so the movement resistances in the individual segments are different. The climbing resistance for a vertical interval of one meter is approximately of the same value as the resistance for a 15 horizontal transportation path of 45 m. In the case shown in the figure, for which the mass distribution is respectively illustrated in the form of four loading blocks, in the upper case the largest mass fraction is transported downhill, whereas in the lower case the 20 largest mass fraction is transported uphill. Up to date, a sliding mean for the average load of the complete belt system under consideration has been formed from the signal of the belt weigher. However, as 25 likewise already mentioned it is not possible to produce a unique functional relationship between the mass flow and the drive power. According to the invention, the procedure now is such 30 that the temporally variable load is assigned to the segments of the belt system that differ with reference to power requirement. The case adapted here in this case is one in which the discharged mass flow is continuously measured via a belt weigher on the belt 35 feeding the belt to be considered. The electric power is measured at the belt to be considered via the power consumption of the motors. Of course, the belt weigher can also be provided on the belt to be considered.
- 7 Every two seconds, the belt weigher supplies a measured value that is intended to constitute a mean value of the load over 2 seconds. The belt speed of all the belts is v=7.5 m/s. In order to be able to assign the 5 measured mass flows m'L,i to the individual segments of the band system, the band system is subdivided into segments i of length 15 m, in addition to which the mean gradient 8i of the belt profile is determined in each case. 10 The measured values of the mass flows must be synchronized with the segments, that is to say for each segment i there belong in relation to the drive power Po measured at the instant to the mass flows m'L,i with that 15 time offset which is required by the load in order to cover the distance between belt weigher and the corresponding segment i. In each clock cycle At, the theoretical power 20 requirement is determined for all segments of the belt system with the currently associated load fractions, and is summed to form a theoretical total power requirement. 25 This first calculation clearly first of all requires the assumption of an estimated fictional coefficient of friction or specific belt resistance f. By making a comparison with the actual measured drive power and variation, it is possible to optimize the specific 30 resistance of the belt system (fictional coefficient of friction) to the effect that the deviations are minimized. This optimization is performed over a representative time period that covers the greatest portion of the occurring load fluctuation. The target 35 function is the errors square sum of the deviations between measured and calculated power over all time increments. The value range for the fictional coefficient of - 8 friction or specific movement resistance fluctuates between 0.010 for very well aligned systems and 0.040 for poorly aligned belt systems. 5 The initially estimated specific resistance is used to calculate the primary resistance FH,i of each subsegment of the n subsegments of the belt system, specifically in the following way: 10 F,i=li-fi-g- [m'R,i + (m'G + m'L,i) -cOSS] li denoting the length of the segment, fi the specific movement resistance, g the acceleration due to gravity, m'R,i the mass of the belt rollers of the relevant 15 segment, m'G the mass of the belt itself, m'L,i the mass of the load, and 8 the mean gradient of the belt profile. The power required per segment is calculated thus: 20 P=v- (FH+FN+FB+FST+Fs) /71 FN denoting the secondary resistances, FE the acceleration resistances, FST the cimbing resistances, 25 and Fs the special resistances, and also q the efficiency of the drive or drives. The acceleration resistance FB occurs only in the region of the transfer, and is calculated as 30 FB = v-m'L/dt and the climbing resistance FST is calculated as 35 FST = g-m'L-sin( 6 ). Secondary resistances FN occur generally only as chute friction in the region of rear station and head station, specifically only above a limit load - 9 determined by design. The best theoretical approach so far to describing the chute friction runs as follows FN = 0 for m'L < =m'limit or 5 FN = k (M'L-m' limit) for m'L > m'iimit The limit value of the load starting from which chute friction occurs is approximately 18000 t/h in the case of the B2.800er belt systems (2800 mm width) . The load 10 independent portion of the primary resistance - that is to say 1-f-g (m'R+m'G*COS(S)) - can be combined over all subsegments and corresponds (when multiplied by v/Tn) to the no load power Pleer. 15 The load dependent portion of the primary resistance and the likewise load dependent climbing resistance can be combined for each segment i to form the following term Fi: 20 Fi=m'L,i'g' (f'cos(Si)+sin(Si)) The total theoretical drive power is therefore Pges= [EFi+F+FN,head+FN, rear] 'v/Tl+Pieer 25 The above-mentioned parameters via which the optimization is performed are as follows: 1. f fictional coefficient of friction 30 2. Pleer No load power 3 m'limit, head Limit load from which chute friction occurs in the head segment 4 head Proportionality factor for the relationship between load and chute 35 friction at the head above the limit load 5 m'limit, rear Limit load from which chute friction occurs in the rear segment 6 krear Proportionality factor for the - 10 relationship between load and chute friction at the rear above the limit load. 5 The f value and the no load power have by far the greatest influence, the no load power mostly being the movement resistance of the dead masses (belt, support rollers) and thus also dependent on the f value. The factors 3 to 6 once again improve the quality of the 10 simulation, but can be neglected if appropriate. The inventive method has the aim of being able to accurately represent the parameters influencing the power requirement via the assignment of the measured 15 mass flow to the individual segments of a belt system and the measured drive power. To this end, the theoretical power requirement is calculated for each individual segment at each instant for which a pair of measured values from drive power and mass flow are 20 present, specifically initially with the aid of estimated values for these parameters. The sum of these theoretical segment powers gives for each instant the theoretical drive power that can be compared with the measured drive power. The sum of the errors squares is 25 a measure of the quality of the estimated values, and is used for an iterative optimization of these estimated values.